封面:芯片战争,克里斯·米勒

CHIP WAR获得更多赞誉

More Praise for CHIP WAR

“多年来我读过的最重要的书之一——引人入胜,文笔优美。米勒表明,尽管存在所有明显的缺陷和失败,但美国资本主义制度一再优于其他制度,并在此过程中为加强民主安全做出了很大贡献。”

“One of the most important books I’ve read in years—engrossing, beautifully written. Miller shows that, for all its manifest flaws and failures, the American capitalist system has repeatedly outperformed other systems and in the process has done much to bolster the security of democracy.”

——罗伯特·卡根,布鲁金斯学会高级研究员, 《华盛顿邮报》专栏作家,《丛林重生:美国和我们危险的世界》的作者

—Robert Kagan, senior fellow, The Brookings Institution, columnist for The Washington Post, and author of The Jungle Grows Back: America and Our Imperiled World

“如果你关心技术,或者美国未来的繁荣,或者它的持续安全,这是一本你必须阅读的书。”

“If you care about technology, or America’s future prosperity, or its continuing security, this is a book you have to read.”

——劳伦斯·H·萨默斯,美国第 71 任财政部长和哈佛大学查尔斯·W·艾略特大学教授

—Lawrence H. Summers, 71st U.S. Secretary of the Treasury and Charles W. Eliot University Professor at Harvard University

“杰出的。米勒的芯片历史涵盖了各个方面:技术、金融,尤其是政治…… 当今最重要的行业之一的首选参考。”

“Outstanding. Miller’s history of the chip covers all angles: technological, financial, and especially political…. The go-to reference on one of the most important industries today.”

——Gavekal Dragonomics 技术分析师 Dan Wang

—Dan Wang, technology analyst at Gavekal Dragonomics

“半导体霸权之争是地缘政治、国家安全和经济繁荣中最重要的故事之一。但它也是最不为人所知的一种。值得庆幸的是,我们现在有了Chip War,让我们对这个重要主题有一个清晰的认识和敏锐的阅读。”

“The battle for supremacy in semiconductors is one of the most important stories in geopolitics, national security, and economic prosperity. But it’s also been one of the least well understood. Thankfully, we now have Chip War to give us a clear view and sharp read on this essential subject.”

——Andrew McAfee,《第二个机器时代》的合著者,《极客之路》《更多来自 Less 》的作者

—Andrew McAfee, coauthor of The Second Machine Age and author of The Geek Way and More from Less

芯片战争,克里斯·米勒,斯克里布纳

致莉亚

To Liya

演员阵容

Cast of Characters

张忠谋:全球最重要的芯片制造商台积电 (TSMC) 创始人;此前,德州仪器公司的高级管理人员。

Morris Chang: Founder of Taiwan Semiconductor Manufacturing Company (TSMC), the world’s most important chipmaker; previously, a senior executive at Texas Instruments.

Andy Grove:1980 年代和 1990 年代英特尔的前总裁兼首席执行官;因其激进的风格和成功重振英特尔而臭名昭著;《只有偏执狂才能生存》的作者。

Andy Grove: Former president and CEO of Intel during the 1980s and 1990s; notorious for his aggressive style and success in reviving Intel; author of Only the Paranoid Survive.

Pat Haggerty:德州仪器董事长;领导该公司,因为它专门制造微电子产品,包括为美国军方制造。

Pat Haggerty: Chairman of Texas Instruments; led the company as it specialized in building microelectronics, including for the U.S. military.

Jack Kilby:集成电路的共同发明者,1958 年;德州仪器的长期雇员;诺贝尔奖得主。

Jack Kilby: Co-inventor of the integrated circuit, in 1958; longtime Texas Instruments employee; winner of the Nobel Prize.

Jay Lathrop:光刻技术的共同发明者,该技术使用专门的化学物质和光对晶体管进行图案化;前身是德州仪器。

Jay Lathrop: Co-inventor of photolithography, the process of patterning transistors using specialized chemicals and light; formerly of Texas Instruments.

Carver Mead:加州理工学院(Caltech)教授;Fairchild Semiconductor 和 Intel 的顾问;对技术的未来有远见的思想家。

Carver Mead: Professor at the California Institute of Technology (Caltech); advisor to Fairchild Semiconductor and Intel; visionary thinker about the future of technology.

Gordon Moore:飞兆半导体和英特尔的联合创始人;1965 年“摩尔定律”的创造者,该定律预测每个芯片上的计算能力每隔几年就会翻一番。

Gordon Moore: Cofounder of Fairchild Semiconductor and Intel; creator in 1965 of “Moore’s Law,” which predicted that the computing power on each chip would double every couple of years.

Akio Morita:索尼联合创始人;《可以说不的日本》的合著者;在 1970 年代和 1980 年代代表日本企业在世界舞台上的表现。

Akio Morita: Cofounder of Sony; coauthor of The Japan That Can Say No; represented Japanese business on the world stage during the 1970s and 1980s.

Robert Noyce:飞兆半导体和英特尔的联合创始人;1959 年集成电路的共同发明人;被誉为“硅谷市长”;Sematech的第一任领导人。

Robert Noyce: Cofounder of Fairchild Semiconductor and Intel; co-inventor of the integrated circuit in 1959; known as the “Mayor of Silicon Valley”; first leader of Sematech.

威廉佩里:五角大楼官员,1977-1981 年,后来的国防部长,1994-1997 年,主张使用芯片生产精确打击武器。

William Perry: Pentagon official from 1977–1981 and later Secretary of Defense from 1994 to 1997 who advocated using chips to produce precision-strike weapons.

Jerry Sanders:AMD 创始人兼 CEO;硅谷最耀眼的推销员;对他认为 1980 年代日本不公平贸易做法的激进批评者。

Jerry Sanders: Founder and CEO of AMD; Silicon Valley’s most flamboyant salesman; an aggressive critic of what he saw as unfair Japanese trade practices in the 1980s.

Charlie Sporck:在飞兆半导体领导制造业务的同时,推动了芯片组装的离岸外包;后任美国国家半导体首席执行官。

Charlie Sporck: Drove the offshoring of chip assembly while leading manufacturing operations at Fairchild Semiconductor; later CEO of National Semiconductor.

任正非:华为创始人,中国电信和芯片设计巨头;他的女儿孟晚舟于2018年在加拿大被捕,罪名是违反美国法律和试图逃避美国制裁。

Ren Zhengfei: Founder of Huawei, China’s telecom and chip-design giant; his daughter Meng Wanzhou was arrested in Canada in 2018 on charges of violating U.S. law and trying to evade U.S. sanctions.

词汇表

Glossary

手臂
一家英国公司,向芯片设计人员授权使用指令集架构——一套管理给定芯片如何运行的基本规则。Arm 架构在移动设备中占主导地位,并正在慢慢赢得 PC 和数据中心的市场份额。
芯片(也称为“集成电路”或“半导体”)
一小块半导体材料,通常是硅,上面刻有数百万或数十亿个微型晶体管。
中央处理器
中央处理器; 一种“通用”芯片,是 PC、手机和数据中心计算的主力。
内存
动态随机存取存储器;两种主要类型的存储芯片之一,用于临时存储数据。
EDA
电子设计自动化;用于设计如何将数百万或数十亿晶体管排列在芯片上并模拟它们的操作的专用软件。
鳍式场效应管
一种新的 3D 晶体管结构在 2010 年代初首次实施,以更好地控制晶体管的操作,因为晶体管的尺寸缩小到纳米级。
显卡
图形处理单元;一种能够并行处理的芯片,使其可用于图形和人工智能应用程序。
逻辑芯片
处理数据的芯片。
内存芯片
记忆数据的芯片。
与非
也称为“闪存”,是第二种主要类型的存储芯片,用于长期数据存储。
光刻
也称为“光刻”;通过图案化掩模照射光或紫外线的过程:光与光刻胶化学物质相互作用,在硅晶片上雕刻图案。
RISC-V
与 Arm 和 x86 不同,一种开源架构越来越受欢迎,因为它可以免费使用。RISC-V 的开发部分由美国政府资助,但现在在中国很受欢迎,因为它不受美国出口管制。
硅片
一块圆形的超纯硅,直径通常为 8 或 12 英寸,上面刻有芯片。
晶体管
一个微小的电子“开关”,可以打开(产生 1)或关闭(0),产生支撑所有数字计算的 1 和 0。
x86
在 PC 和数据中心中占主导地位的指令集架构。英特尔和 AMD 是生产此类芯片的两家主要公司。

介绍

Introduction

2020年 8 月 18 日,美国海军陆战队驱逐舰 USS Mustin滑入台湾海峡北端,其 5 英寸舰炮指向南方,开始执行单独穿越海峡的任务,并重申这些国际水域并非由中国控制——至少现在还没有。一股强烈的西南风吹过甲板,向南飞去。高高的云层在水面上投下阴影,似乎一直延伸到福州、厦门、香港等港口城市,以及散布在华南海岸的其他港口。东边,台湾岛在远处升起,一片广阔而密集的沿海平原,让位于隐藏在云层中的高耸山峰。船上,一名戴着海军棒球帽、戴着外科口罩的水手举起双筒望远镜扫视着地平线。水域充满了商业货轮,将货物从亚洲的工厂运送到世界各地的消费者。

The destroyer USS Mustin slipped into the northern end of the Taiwan Strait on August 18, 2020, its five-inch gun pointed southward as it began a solo mission to sail through the Strait and reaffirm that these international waters were not controlled by China—at least not yet. A stiff southwestern breeze whipped across the deck as it steamed south. High clouds cast shadows on the water that seemed to stretch all the way to the great port cities of Fuzhou, Xiamen, Hong Kong, and the other harbors that dot the South China coast. To the east, the island of Taiwan rose in the distance, a broad, densely settled coastal plain giving way to tall peaks hidden in clouds. Aboard ship, a sailor wearing a navy baseball cap and a surgical mask lifted his binoculars and scanned the horizon. The waters were filled with commercial freighters shipping goods from Asia’s factories to consumers around the world.

在 USS Mustin上,一排水手坐在一个黑暗的房间里,前面是一排色彩鲜艳的屏幕,上面显示着来自飞机、无人机、船只和卫星的数据,这些数据跟踪了印度-太平洋的运动。在Mustin的舰桥上,一个雷达阵列被输入到船上的计算机中。甲板上的九十六个发射单元准备就绪,每个发射单元都能发射导弹,可以精确打击数十甚至数百英里外的飞机、船只或潜艇。危机期间冷战时期,美军曾以强力核武威胁保卫台湾。今天,它依赖于微电子和精确打击。

On board the USS Mustin, a row of sailors sat in a dark room in front of an array of brightly colored screens on which were displayed data from planes, drones, ships, and satellites tracking movement across the Indo-Pacific. Atop the Mustin’s bridge, a radar array fed into the ship’s computers. On deck ninety-six launch cells stood ready, each capable of firing missiles that could precisely strike planes, ships, or submarines dozens or even hundreds of miles away. During the crises of the Cold War, the U.S. military had used threats of brute nuclear force to defend Taiwan. Today, it relies on microelectronics and precision strikes.

当美国海军陆战队满载电脑化武器驶过海峡时,中国人民解放军宣布在台湾进行一系列报复性实弹演习,实施北京控制的报纸所说的“武力统一行动。” 但在这一天,中国领导人对美国海军的担忧较少,而对美国商务部一项名为“实体清单”的晦涩难懂的法规感到担忧,该法规限制了美国技术向海外的转移。此前,实体清单主要用于阻止导弹部件或核材料等军事系统的销售。不过,现在美国政府正在大幅收紧管理计算机芯片的规则,这些规则已在军事系统和消费品中无处不在。

As the USS Mustin sailed through the Strait, bristling with computerized weaponry, the People’s Liberation Army announced a retaliatory series of live-fire exercises around Taiwan, practicing what one Beijing-controlled newspaper called a “reunification-by-force operation.” But on this particular day, China’s leaders worried less about the U.S. Navy and more about an obscure U.S. Commerce Department regulation called the Entity List, which limits the transfer of American technology abroad. Previously, the Entity List had primarily been used to prevent sales of military systems like missile parts or nuclear materials. Now, though, the U.S. government was dramatically tightening the rules governing computer chips, which had become ubiquitous in both military systems and consumer goods.

目标是中国科技巨头华为,该公司销售智能手机、电信设备、云计算服务和其他先进技术。美国担心华为的产品现在定价如此具有吸引力,部分原因是中国政府的补贴,以至于它们很快就会成为下一代电信网络的支柱。美国在世界科技基础设施中的主导地位将受到削弱。中国的地缘政治影响力将会增长。为了应对这一威胁,美国禁止华为购买采用美国技术制造的先进计算机芯片。

The target was Huawei, China’s tech giant, which sells smartphones, telecom equipment, cloud computing services, and other advanced technologies. The U.S. feared that Huawei’s products were now priced so attractively, partly owing to Chinese government subsidies, that they’d shortly form the backbone of next-generation telecom networks. America’s dominance of the world’s tech infrastructure would be undermined. China’s geopolitical clout would grow. To counter this threat, the U.S. barred Huawei from buying advanced computer chips made with U.S. technology.

很快,该公司的全球扩张陷入停顿。整个产品线变得无法生产。收入下滑。一家企业巨头面临技术窒息。华为发现,与所有其他中国公司一样,它严重依赖外国人来制造所有现代电子产品所依赖的芯片。

Soon, the company’s global expansion ground to a halt. Entire product lines became impossible to produce. Revenue slumped. A corporate giant faced technological asphyxiation. Huawei discovered that, like all other Chinese companies, it was fatally dependent on foreigners to make the chips upon which all modern electronics depend.

美国仍然牢牢控制着让硅谷得名的硅芯片,尽管它的地位已经危险地削弱了。中国现在每年在进口芯片上的花费比在石油上的花费还多。这些半导体以各种方式插入从智能手机到冰箱,中国在国内消费或在全球出口的设备数量。战略家们对中国的“马六甲困境”(指太平洋和印度洋之间的主要航运通道)以及该国在危机期间获得石油和其他商品供应的能力进行了理论分析。然而,北京更担心以字节而不是桶来衡量的封锁。中国正在倾注自己的智慧和数十亿美元开发自己的半导体技术,以期摆脱美国芯片扼流圈。

The United States still has a stranglehold on the silicon chips that gave Silicon Valley its name, though its position has weakened dangerously. China now spends more money each year importing chips than it spends on oil. These semiconductors are plugged into all manner of devices, from smartphones to refrigerators, that China consumes at home or exports worldwide. Armchair strategists theorize about China’s “Malacca Dilemma”—a reference to the main shipping channel between the Pacific and Indian Oceans—and the country’s ability to access supplies of oil and other commodities amid a crisis. Beijing, however, is more worried about a blockade measured in bytes rather than barrels. China is devoting its best minds and billions of dollars to developing its own semiconductor technology in a bid to free itself from America’s chip choke.

如果北京成功,它将重塑全球经济并重新设定军事力量平衡。第二次世界大战由钢和铝决定,紧随其后的是由原子武器定义的冷战。中美之间的竞争很可能是由计算能力决定的。北京和华盛顿的战略家现在意识到,所有先进技术——从机器学习到导弹系统,从自动驾驶汽车到武装无人机——都需要尖端芯片,更正式地称为半导体或集成电路。少数公司控制他们的生产。

If Beijing succeeds, it will remake the global economy and reset the balance of military power. World War II was decided by steel and aluminum, and followed shortly thereafter by the Cold War, which was defined by atomic weapons. The rivalry between the United States and China may well be determined by computing power. Strategists in Beijing and Washington now realize that all advanced tech—from machine learning to missile systems, from automated vehicles to armed drones—requires cutting-edge chips, known more formally as semiconductors or integrated circuits. A tiny number of companies control their production.

我们很少想到芯片,但它们创造了现代世界。国家的命运已经取决于他们驾驭计算能力的能力。我们知道,如果没有半导体贸易及其使电子产品成为可能,全球化就不会存在。美国的军事优势很大程度上源于其将芯片应用于军事用途的能力。亚洲在过去半个世纪的巨大崛起建立在硅的基础上,因为其不断增长的经济体已经开始专注于制造芯片以及组装这些集成电路使之成为可能的计算机和智能手机。

We rarely think about chips, yet they’ve created the modern world. The fate of nations has turned on their ability to harness computing power. Globalization as we know it wouldn’t exist without the trade in semiconductors and the electronic products they make possible. America’s military primacy stems largely from its ability to apply chips to military uses. Asia’s tremendous rise over the past half century has been built on a foundation of silicon as its growing economies have come to specialize in fabricating chips and assembling the computers and smartphones that these integrated circuits make possible.

计算的核心是需要数百万个 1 和 0。整个数字世界由这两个数字组成。iPhone 上的每一个按钮、每封电子邮件、每张照片和 YouTube 视频——所有这些最终都被编码成大量的 1 和 0 字符串。但这些数字实际上并不存在。它们是电流的表达,开启 (1) 或关闭 (0)。芯片是由数百万或数十亿个晶体管组成的网格,这些晶体管是微小的电子开关,可以打开和关闭以处理这些数字,记住它们,并将图像、声音和无线电波等现实世界的感觉转换为数百万个 1 和0 秒。

At the core of computing is the need for many millions of 1s and 0s. The entire digital universe consists of these two numbers. Every button on your iPhone, every email, photograph, and YouTube video—all of these are coded, ultimately, in vast strings of 1s and 0s. But these numbers don’t actually exist. They’re expressions of electrical currents, which are either on (1) or off (0). A chip is a grid of millions or billions of transistors, tiny electrical switches that flip on and off to process these digits, to remember them, and to convert real world sensations like images, sound, and radio waves into millions and millions of 1s and 0s.

随着美国海军陆战队向南航行,海峡两岸的工厂和装配厂都在为 iPhone 12 生产零部件,而距离 2020 年 10 月的发布只有两个月的时间。大约芯片行业四分之一的收入来自手机;一部新手机的大部分价格都用于购买里面的半导体。在过去的十年中,每一代 iPhone 都采用了世界上最先进的处理器芯片之一。总的来说,要让智能手机工作需要十几种半导体,不同的芯片管理电池、蓝牙、Wi-Fi、蜂窝网络连接、音频、相机等。

As the USS Mustin sailed southward, factories and assembly facilities on both sides of the Strait were churning out components for the iPhone 12, which was only two months away from its October 2020 launch. Around a quarter of the chip industry’s revenue comes from phones; much of the price of a new phone pays for the semiconductors inside. For the past decade, each generation of iPhone has been powered by one of the world’s most advanced processor chips. In total, it takes over a dozen semiconductors to make a smartphone work, with different chips managing the battery, Bluetooth, Wi-Fi, cellular network connections, audio, the camera, and more.

苹果完全生产这些芯片。它购买了大多数现成的产品:日本 Kioxia 的存储芯片、加利福尼亚 Skyworks 的射频芯片、德克萨斯州奥斯汀的 Cirrus Logic 的音频芯片。Apple 内部设计了运行 iPhone 操作系统的超复杂处理器。但是,加利福尼亚州库比蒂诺的巨人无法制造这些芯片。美国、欧洲、日本或中国的任何公司也不能。今天,Apple 最先进的处理器——可以说是世界上最先进的半导体——只能由一家公司在一座大楼内生产,即人类历史上最昂贵的工厂,于 2020 年 8 月 18 日上午,距离美国海军陆战队的右舷船头只有几十英里。

Apple makes precisely none of these chips. It buys most off-the-shelf: memory chips from Japan’s Kioxia, radio frequency chips from California’s Skyworks, audio chips from Cirrus Logic, based in Austin, Texas. Apple designs in-house the ultra-complex processors that run an iPhone’s operating system. But the Cupertino, California, colossus can’t manufacture these chips. Nor can any company in the United States, Europe, Japan, or China. Today, Apple’s most advanced processors—which are arguably the world’s most advanced semiconductors—can only be produced by a single company in a single building, the most expensive factory in human history, which on the morning of August 18, 2020, was only a couple dozen miles off the USS Mustin’s starboard bow.

制造和小型化半导体一直是我们这个时代最大的工程挑战。今天,没有哪家公司比台积电(TSMC)制造的芯片精度更高。2020 年,当世界陷入由直径约为 100 纳米(十亿分之一米)的病毒驱动的封锁之间时,台积电最先进的工厂 Fab 18 正在雕刻微型晶体管的微观迷宫,蚀刻的形状小于一种冠状病毒,百分之一线粒体大小。台积电以人类历史上前所未有的规模复制了这一过程。苹果售出超过 1 亿部 iPhone 12,每部都由 A14 处理器芯片驱动,其硅片上刻有 118 亿个微型晶体管。换句话说,在几个月的时间里,台积电的 Fab 18 仅针对 iPhone 中的十几个芯片中的一个,就制造了超过 1 万亿个晶体管——也就是说,这个数字后面有 18 个零。去年,芯片行业生产的晶体管数量超过了人类历史上所有其他行业的所有其他公司生产的所有产品的总和。没有其他东西能接近。

Fabricating and miniaturizing semiconductors has been the greatest engineering challenge of our time. Today, no firm fabricates chips with more precision than the Taiwan Semiconductor Manufacturing Company, better known as TSMC. In 2020, as the world lurched between lockdowns driven by a virus whose diameter measured around one hundred nanometers—billionths of a meter—TSMC’s most advanced facility, Fab 18, was carving microscopic mazes of tiny transistors, etching shapes smaller than half the size of a coronavirus, a hundredth the size of a mitochondria. TSMC replicated this process at a scale previously unparalleled in human history. Apple sold over 100 million iPhone 12s, each powered by an A14 processor chip with 11.8 billion tiny transistors carved into its silicon. In a matter of months, in other words, for just one of the dozen chips in an iPhone, TSMC’s Fab 18 fabricated well over 1 quintillion transistors—that is, a number with eighteen zeros behind it. Last year, the chip industry produced more transistors than the combined quantity of all goods produced by all other companies, in all other industries, in all human history. Nothing else comes close.

仅仅在 60 年前,尖端芯片上的晶体管数量不是 118 亿,而是 4。 1961 年,旧金山南部一家名为仙童半导体的小公司宣布推出一种名为 Micrologic 的新产品,这是一种嵌入了四个晶体管的硅芯片。很快,该公司设计了将十几个晶体管放在一个芯片上的方法,然后是一百个。Fairchild 的联合创始人 Gordon Moore 在 1965 年注意到,随着工程师学会制造越来越小的晶体管,每个芯片上可以安装的组件数量每年翻一番。这一预测——芯片的计算能力将呈指数级增长——被称为“摩尔定律”,并导致摩尔预测了 1965 年设备的发明,这些设备在 1965 年似乎具有不可思议的未来感,例如“电子手表”、“家用电脑”和甚至是“个人便携式通信设备”。展望1965年,摩尔预测了十年的指数增长——但这种惊人的进步速度已经持续了半个多世纪。1970 年,摩尔创立的第二家公司英特尔推出了一种可以记住 1,024 条信息(“比特”)的存储芯片。大概20美元左右每比特两美分。今天,20 美元就可以买到一个记忆力超过 10 亿位的 U 盘。

It was only sixty years ago that the number of transistors on a cutting-edge chip wasn’t 11.8 billion, but 4. In 1961, south of San Francisco, a small firm called Fairchild Semiconductor announced a new product called the Micrologic, a silicon chip with four transistors embedded in it. Soon the company devised ways to put a dozen transistors on a chip, then a hundred. Fairchild cofounder Gordon Moore noticed in 1965 that the number of components that could be fit on each chip was doubling annually as engineers learned to fabricate ever smaller transistors. This prediction—that the computing power of chips would grow exponentially—came to be called “Moore’s Law” and led Moore to predict the invention of devices that in 1965 seemed impossibly futuristic, like an “electronic wristwatch,” “home computers,” and even “personal portable communications equipment.” Looking forward from 1965, Moore predicted a decade of exponential growth—but this staggering rate of progress has continued for over half a century. In 1970, the second company Moore founded, Intel, unveiled a memory chip that could remember 1,024 pieces of information (“bits”). It cost around $20, roughly two cents per bit. Today, $20 can buy a thumb drive that can remember well over a billion bits.

当我们今天想到硅谷时,我们的脑海中会联想到社交网络和软件公司,而不是硅谷命名的材料。然而,互联网、云、社交媒体和整个数字世界的存在只是因为工程师学会了控制电子在硅片上的最微小运动。如果在过去半个世纪中处理和记住 1 和 0 的成本没有下降十亿倍,那么“大科技”就不会存在。

When we think of Silicon Valley today, our minds conjure social networks and software companies rather than the material after which the valley was named. Yet the internet, the cloud, social media, and the entire digital world only exist because engineers have learned to control the most minute movement of electrons as they race across slabs of silicon. “Big tech” wouldn’t exist if the cost of processing and remembering 1s and 0s hadn’t fallen by a billionfold in the past half century.

这一令人难以置信的上升部分归功于杰出的科学家和诺贝尔奖获得者物理学家。但并非每项发明都能创造出成功的创业公司,也并非每项创业公司都能激发出改变世界的新行业。半导体之所以在社会中传播开来,是因为公司发明了新的技术来制造数以百万计的半导体,因为强硬的管理人员不断地降低成本,也因为富有创造力的企业家想出了使用它们的新方法。摩尔定律的形成既是关于制造专家、供应链专家和营销经理的故事,也是关于物理学家或电气工程师的故事。

This incredible ascent is partly thanks to brilliant scientists and Nobel Prize−winning physicists. But not every invention creates a successful startup, and not every startup sparks a new industry that transforms the world. Semiconductors spread across society because companies devised new techniques to manufacture them by the millions, because hard-charging managers relentlessly drove down their cost, and because creative entrepreneurs imagined new ways to use them. The making of Moore’s Law is as much a story of manufacturing experts, supply chain specialists, and marketing managers as it is about physicists or electrical engineers.

旧金山南部的城镇——直到 1970 年代才被称为硅谷——是这场革命的中心,因为它们结合了科学专业知识、制造技术和富有远见的商业思维。加州有很多从斯坦福或伯克利毕业的航空或无线电行业的工程师,在美国军方试图巩固其技术优势时,每个人都拥有大量国防资金。然而,加州的文化与任何经济结构一样重要。离开美国东海岸、欧洲和亚洲建立芯片产业的人,在决定搬到硅谷时,常常以无限机遇感为由。对于世界上最聪明的工程师和最具创造力的企业家来说,没有比这更令人兴奋的地方了。

The towns to the south of San Francisco—which weren’t called Silicon Valley until the 1970s—were the epicenter of this revolution because they combined scientific expertise, manufacturing know-how, and visionary business thinking. California had plenty of engineers trained in aviation or radio industries who’d graduated from Stanford or Berkeley, each of which was flush with defense dollars as the U.S. military sought to solidify its technological advantage. California’s culture mattered just as much as any economic structure, however. The people who left America’s East Coast, Europe, and Asia to build the chip industry often cited a sense of boundless opportunity in their decision to move to Silicon Valley. For the world’s smartest engineers and most creative entrepreneurs, there was simply no more exciting place to be.

芯片产业一旦成型,就不可能脱离硅谷。今天的半导体供应链需要来自许多城市和国家的组件,但几乎每个制造的芯片都与硅谷有联系,或者是使用在加利福尼亚设计和制造的工具生产的。美国庞大的科学专业知识储备,得到政府研究资金的培育,并通过从其他国家挖来最优秀科学家的能力得到加强,为核心知识推动技术进步。该国的风险投资公司网络及其股票市场为新公司提供了发展所需的启动资金,并无情地淘汰了倒闭的公司。与此同时,全球最大的消费市场美国推动了数十年新型芯片研发的增长。

Once the chip industry took shape, it proved impossible to dislodge from Silicon Valley. Today’s semiconductor supply chain requires components from many cities and countries, but almost every chip made still has a Silicon Valley connection or is produced with tools designed and built in California. America’s vast reserve of scientific expertise, nurtured by government research funding and strengthened by the ability to poach the best scientists from other countries, has provided the core knowledge driving technological advances forward. The country’s network of venture capital firms and its stock markets have provided the startup capital new firms need to grow—and have ruthlessly forced out failing companies. Meanwhile, the world’s largest consumer market in the U.S. has driven the growth that’s funded decades of R&D on new types of chips.

其他国家发现自己无法跟上步伐,但当它们深度融入硅谷的供应链时,它们就取得了成功。欧洲拥有孤立的半导体专业知识岛,特别是在生产制造芯片所需的机床和设计芯片架构方面。台湾、韩国和日本的亚洲政府通过补贴公司、资助培训项目、保持汇率低估以及对进口芯片征收关税等方式,挤进了芯片行业。这一战略产生了其他国家无法复制的某些能力——但他们已经取得了与硅谷合作的成果,继续从根本上依赖美国的工具、软件和客户。同时,

Other countries have found it impossible to keep up on their own but have succeeded when they’ve deeply integrated themselves into Silicon Valley’s supply chains. Europe has isolated islands of semiconductor expertise, notably in producing the machine tools needed to make chips and in designing chip architectures. Asian governments, in Taiwan, South Korea, and Japan, have elbowed their way into the chip industry by subsidizing firms, funding training programs, keeping their exchange rates undervalued, and imposing tariffs on imported chips. This strategy has yielded certain capabilities that no other countries can replicate—but they’ve achieved what they have in partnership with Silicon Valley, continuing to rely fundamentally on U.S. tools, software, and customers. Meanwhile, America’s most successful chip firms have built supply chains that stretch across the world, driving down costs and producing the expertise that has made Moore’s Law possible.

今天,由于摩尔定律,半导体被嵌入到每一个需要计算能力的设备中——在物联网时代,这意味着几乎每一个设备。即使是像汽车这样的百年产品,现在也经常包含价值一千美元的芯片。世界上大部分的 GDP 都是由依赖于半导体的设备产生的。对于 75 年前还不存在的产品来说,这是一次非凡的提升。

Today, thanks to Moore’s Law, semiconductors are embedded in every device that requires computing power—and in the age of the Internet of Things, this means pretty much every device. Even hundred-year-old products like automobiles now often include a thousand dollars worth of chips. Most of the world’s GDP is produced with devices that rely on semiconductors. For a product that didn’t exist seventy-five years ago, this is an extraordinary ascent.

随着美国海军陆战队在 2020 年 8 月向南驶去,世界才刚刚开始考虑我们对半导体的依赖——以及我们对制造芯片的台湾的依赖。我们每年使用的新计算能力的三分之一。台湾台积电建几乎所有世界上最先进的处理器芯片。当 COVID 在 2020 年席卷全球时,它扰乱了芯片行业,也。一些工厂暂时关闭。汽车芯片的采购量下降。随着世界上许多人准备在家工作,对个人电脑和数据中心芯片的需求飙升。然后,在 2021 年,发生了一系列事故——日本半导体工厂发生火灾;美国芯片制造中心德克萨斯州的冰风暴;以及在马来西亚组装和测试许多芯片的新一轮 COVID 封锁加剧了这些中断。突然之间,许多远离硅谷的行业都面临着严重的芯片短缺。从丰田到大型汽车制造商通用汽车不得不关闭工厂数周,因为他们无法获得所需的半导体。即使是最简单的芯片也出现短缺,导致世界另一端的工厂关闭。这似乎是一个完美的全球化形象出了问题。

As the USS Mustin steamed southward in August 2020, the world was just beginning to reckon with our reliance on semiconductors—and our dependence on Taiwan, which fabricates the chips that produce a third of the new computing power we use each year. Taiwan’s TSMC builds almost all the world’s most advanced processor chips. When COVID slammed into the world in 2020, it disrupted the chip industry, too. Some factories were temporarily shuttered. Purchases of chips for autos slumped. Demand for PC and data center chips spiked higher, as much of the world prepared to work from home. Then, over 2021, a series of accidents—a fire in a Japanese semiconductor facility; ice storms in Texas, a center of U.S. chipmaking; and a new round of COVID lockdowns in Malaysia, where many chips are assembled and tested—intensified these disruptions. Suddenly, many industries far from Silicon Valley faced debilitating chip shortages. Big carmakers from Toyota to General Motors had to shut factories for weeks because they couldn’t acquire the semiconductors they needed. Shortages of even the simplest chips caused factory closures on the opposite side of the world. It seemed like a perfect image of globalization gone wrong.

几十年来,美国、欧洲和日本的政治领导人对半导体的考虑并不多。和我们其他人一样,他们认为“技术”是指搜索引擎或社交媒体,而不是硅晶片。当乔·拜登和安吉拉·默克尔问到为什么他们国家的汽车工厂被关闭时,答案隐藏在令人眼花缭乱的半导体供应链背后。一个典型的芯片可能由位于英国的日资公司 Arm 的蓝图设计,由加利福尼亚和以色列的工程师团队使用美国的设计软件设计。当设计完成后,它会被送到台湾的一家工厂,该工厂从日本购买超纯硅晶片和特殊气体。该设计是使用世界上最精密的机器雕刻在硅上的,这些机器可以蚀刻、沉积、并测量几原子厚的材料层。这些工具主要由 5 家公司生产,1 家荷兰人、1 家日本公司和 3 家加利福尼亚人,没有它们,高级芯片基本上是不可能制造出来的。然后,通常在东南亚对芯片进行封装和测试,然后再送往中国组装成手机或电脑。

Political leaders in the U.S., Europe, and Japan hadn’t thought much about semiconductors in decades. Like the rest of us, they thought “tech” meant search engines or social media, not silicon wafers. When Joe Biden and Angela Merkel asked why their country’s car factories were shuttered, the answer was shrouded behind semiconductor supply chains of bewildering complexity. A typical chip might be designed with blueprints from the Japanese-owned, UK-based company called Arm, by a team of engineers in California and Israel, using design software from the United States. When a design is complete, it’s sent to a facility in Taiwan, which buys ultra-pure silicon wafers and specialized gases from Japan. The design is carved into silicon using some of the world’s most precise machinery, which can etch, deposit, and measure layers of materials a few atoms thick. These tools are produced primarily by five companies, one Dutch, one Japanese, and three Californian, without which advanced chips are basically impossible to make. Then the chip is packaged and tested, often in Southeast Asia, before being sent to China for assembly into a phone or computer.

如果半导体生产过程中的任何一个步骤被中断,全球新计算能力的供应就会受到威胁。在人工智能时代,人们常说数据是新的石油。然而,我们面临的真正限制不是数据的可用性,而是处理能力。可以存储和处理数据的半导体数量有限。生产它们是令人难以置信的复杂和可怕的昂贵。与可以从许多国家购买的石油不同,我们的计算能力生产从根本上依赖于一系列瓶颈:工具、化学品和软件,它们通常由少数几家公司生产——有时仅由一家公司生产。没有其他经济方面如此依赖如此少的公司。来自台湾的芯片每年提供全球 37% 的新计算能力。两家韩国公司生产全球44%的内存芯片。荷兰公司 ASML 制造了世界上 100% 的极紫外光刻机,没有它根本不可能制造出尖端芯片。相比之下,欧佩克占世界石油产量 40% 的份额显得微不足道。

If any one of the steps in the semiconductor production process is interrupted, the world’s supply of new computing power is imperiled. In the age of AI, it’s often said that data is the new oil. Yet the real limitation we face isn’t the availability of data but of processing power. There’s a finite number of semiconductors that can store and process data. Producing them is mind-bogglingly complex and horrendously expensive. Unlike oil, which can be bought from many countries, our production of computing power depends fundamentally on a series of choke points: tools, chemicals, and software that often are produced by a handful of companies—and sometimes only by one. No other facet of the economy is so dependent on so few firms. Chips from Taiwan provide 37 percent of the world’s new computing power each year. Two Korean companies produce 44 percent of the world’s memory chips. The Dutch company ASML builds 100 percent of the world’s extreme ultraviolet lithography machines, without which cutting-edge chips are simply impossible to make. OPEC’s 40 percent share of world oil production looks unimpressive by comparison.

每年生产一万亿纳米级芯片的全球公司网络是效率的胜利。这也是一个惊人的漏洞。大流行的破坏只是让我们瞥见了一次恰到好处的地震可能对全球经济造成的影响。台湾位于断层线之上,该断层线最近在 1999 年发生了里氏 7.3 级地震。值得庆幸的是,这只让芯片生产停工了几天。但更强烈的地震袭击台湾只是时间问题。一场毁灭性的地震也可能袭击日本,一个地震多发的国家,生产全球 17% 的芯片,或者硅谷,今天生产的芯片很少,但在圣安德烈亚斯断层顶部的设施中建造关键的芯片制造机器。

The global network of companies that annually produces a trillion chips at nanometer scale is a triumph of efficiency. It’s also a staggering vulnerability. The disruptions of the pandemic provide just a glimpse of what a single well-placed earthquake could do to the global economy. Taiwan sits atop a fault line that as recently as 1999 produced an earthquake measuring 7.3 on the Richter scale. Thankfully, this only knocked chip production offline for a couple of days. But it’s only a matter of time before a stronger quake strikes Taiwan. A devastating quake could also hit Japan, an earthquake-prone country that produces 17 percent of the world’s chips, or Silicon Valley, which today produces few chips but builds crucial chipmaking machinery in facilities sitting atop the San Andreas Fault.

然而,当今最危及半导体供应的巨变不是构造板块的崩溃,而是大国的冲突。随着中国和美国争夺霸权,华盛顿和北京都专注于控制计算的未来——而且,在一个可怕的程度上,这个未来取决于一个被北京视为叛离省份而美国承诺保卫的小岛通过武力。

Yet the seismic shift that most imperils semiconductor supply today isn’t the crash of tectonic plates but the clash of great powers. As China and the United States struggle for supremacy, both Washington and Beijing are fixated on controlling the future of computing—and, to a frightening degree, that future is dependent on a small island that Beijing considers a renegade province and America has committed to defend by force.

中美芯片产业之间的互联互通台湾的情况复杂得令人眼花缭乱。没有比创立台积电更好的例证了,台积电直到 2020 年都将美国的苹果和中国的华为视为其最大的两个客户。张忠谋出生于中国大陆;在二战时期的香港长大;曾在哈佛大学、麻省理工学院和斯坦福大学接受教育;在达拉斯的德州仪器工作期间帮助建立了美国早期的芯片产业;举行了一个绝密的美国为美国军方开发电子产品的安全许可;使台湾成为世界半导体制造的中心。北京和华盛顿的一些外交政策战略家梦想着将两国的科技行业脱钩,但像张这样的人帮助建立的由芯片设计师、化学品供应商和机床制造商组成的超高效国际网络并不能轻易解散。

The interconnections between the chip industries in the U.S., China, and Taiwan are dizzyingly complex. There’s no better illustration of this than the individual who founded TSMC, a company that until 2020 counted America’s Apple and China’s Huawei as its two biggest customers. Morris Chang was born in mainland China; grew up in World War II−era Hong Kong; was educated at Harvard, MIT, and Stanford; helped build America’s early chip industry while working for Texas Instruments in Dallas; held a top secret U.S. security clearance to develop electronics for the American military; and made Taiwan the epicenter of world semiconductor manufacturing. Some foreign policy strategists in Beijing and Washington dream of decoupling the two countries’ tech sectors, but the ultra-efficient international network of chip designers, chemical suppliers, and machine-tool makers that people like Chang helped build can’t be easily unwound.

当然,除非发生爆炸。北京明确拒绝排除可能入侵台湾以与大陆“统一”的可能性。但是,要让半导体引发的冲击波在全球经济中肆虐,并不需要像两栖攻击那样戏剧性的事情。即使是中国军队的部分封锁也会引发毁灭性的破坏。一旦电话、数据中心、汽车、电信网络和其他技术的生产延迟加起来,对台积电最先进的芯片制造设施的一次导弹袭击很容易造成数千亿美元的损失。

Unless, of course, something explodes. Beijing has pointedly refused to rule out the prospect that it might invade Taiwan to “reunify” it with the mainland. But it wouldn’t take anything as dramatic as an amphibious assault to send semiconductor-induced shock waves careening through the global economy. Even a partial blockade by Chinese forces would trigger devastating disruptions. A single missile strike on TSMC’s most advanced chip fabrication facility could easily cause hundreds of billions of dollars of damage once delays to the production of phones, data centers, autos, telecom networks, and other technology are added up.

让全球经济成为世界上最危险的政治争端之一的人质似乎是一个具有历史意义的错误。然而,先进芯片制造业集中在台湾、韩国和东亚其他地区并非偶然。政府官员和企业高管的一系列深思熟虑的决定创造了我们今天所依赖的广泛供应链。亚洲大量廉价劳动力吸引了寻找低成本工厂工人的芯片制造商。该地区的政府和企业使用离岸芯片组装设施来了解并最终驯化更先进的技术。华盛顿的外交政策战略家采用复杂的半导体供应链作为将亚洲与美国主导的世界联系起来的工具。资本主义对经济效率的无情需求推动了成本削减和企业整合的不断推动。支持摩尔定律的技术创新的稳定节奏需要更复杂的材料、机械和工艺,而这些材料、机械和工艺只能通过全球市场提供或资助。我们对计算能力的巨大需求只会继续增长。

Holding the global economy hostage to one of the world’s most dangerous political disputes might seem like an error of historic proportions. However, the concentration of advanced chip manufacturing in Taiwan, South Korea, and elsewhere in East Asia isn’t an accident. A series of deliberate decisions by government officials and corporate executives created the far-flung supply chains we rely on today. Asia’s vast pool of cheap labor attracted chipmakers looking for low-cost factory workers. The region’s governments and corporations used offshored chip assembly facilities to learn about, and eventually domesticate, more advanced technologies. Washington’s foreign policy strategists embraced complex semiconductor supply chains as a tool to bind Asia to an American-led world. Capitalism’s inexorable demand for economic efficiency drove a constant push for cost cuts and corporate consolidation. The steady tempo of technological innovation that underwrote Moore’s Law required ever more complex materials, machinery, and processes that could only be supplied or funded via global markets. And our gargantuan demand for computing power only continues to grow.

这本书借鉴了从台北到莫斯科三大洲的历史档案研究,以及对科学家、工程师、首席执行官和政府官员的一百多次采访,认为半导体定义了我们生活的世界,决定了国际政治的形态、世界经济的结构和军事力量的平衡。然而,这种最现代的设备有着复杂而充满争议的历史。它的发展不仅受到企业和消费者的影响,还受到雄心勃勃的政府和战争的必要性的影响。要了解我们的世界是如何由数以十亿计的晶体管和少数不可替代的公司定义的,我们必须从回顾硅时代的起源开始。

Drawing on research in historical archives on three continents, from Taipei to Moscow, and over a hundred interviews with scientists, engineers, CEOs, and government officials, this book contends that semiconductors have defined the world we live in, determining the shape of international politics, the structure of the world economy, and the balance of military power. Yet this most modern of devices has a complex and contested history. Its development has been shaped not only by corporations and consumers but also by ambitious governments and the imperatives of war. To understand how our world came to be defined by quintillions of transistors and a tiny number of irreplaceable companies, we must begin by looking back to the origins of the silicon age.

第一部分 冷战筹码

PART I COLD WAR CHIPS

第1章 从钢铁到硅

CHAPTER 1 From Steel to Silicon

日本士兵将二战形容为“钢铁台风”。对森田昭夫来说,确实有这种感觉,出身于清酒商人家庭,勤奋好学的年轻工程师。森田只是勉强避开了前线,被分配到日本海军工程实验室。但钢铁台风也袭击了森田的家乡,美国 B-29 超级堡垒轰炸机轰炸日本城市,摧毁了东京和其他城市中心的大部分地区。雪上加霜的是,美国的封锁造成了广泛的饥饿,迫使该国采取了绝望的措施。战争结束时,森田的兄弟们正在接受神风特攻队飞行员的训练。

Japanese soldiers described World War II as a “typhoon of steel.” It certainly felt that way to Akio Morita, a studious young engineer from a family of prosperous sake merchants. Morita only barely avoided the front lines by getting assigned to a Japanese navy engineering lab. But the typhoon of steel crashed through Morita’s homeland, too, as American B-29 Superfortress bombers pummeled Japan’s cities, destroying much of Tokyo and other urban centers. Adding to the devastation, an American blockade created widespread hunger and drove the country toward desperate measures. Morita’s brothers were being trained as kamikaze pilots when the war ended.

横跨东海,莫里斯·张的童年被枪声和空袭警报警告即将袭击的声音打断。少年时期,常青逃离横扫中国的日军,移居广州。香港的英国殖民地;中国战时首都重庆;日军战败后返回上海。即便如此,战争并没有真正结束,因为共产党游击队重新开始了与中国政府的斗争。不久,毛泽东的军队向上海进发。张忠谋再次成为难民,第二次被迫逃往香港。

Across the East China Sea, Morris Chang’s childhood was punctuated by the sound of gunfire and air-raid sirens warning of imminent attack. Chang spent his teenage years fleeing the Japanese armies that swept across China, moving to Guangzhou; the British colony of Hong Kong; China’s wartime capital of Chongqing; and then back to Shanghai after the Japanese were defeated. Even then, the war didn’t really end, because Communist guerillas relaunched their struggle against the Chinese government. Soon Mao Zedong’s forces were marching on Shanghai. Morris Chang was once again a refugee, forced to flee to Hong Kong for the second time.

布达佩斯在世界的另一边,但是安迪·格鲁夫(Andy Grove)经历了席卷亚洲的钢铁台风。安迪(或安德拉斯格罗夫,他当时的名字)在布达佩斯的多次入侵中幸存下来。匈牙利极右翼政府将犹太人视为二等公民,但当欧洲爆发战争时,他的父亲仍然被征召入伍,与匈牙利的纳粹盟友一起对抗苏联,据报道他在行动中失踪斯大林格勒。然后,在 1944 年,纳粹入侵了他们表面上的盟友匈牙利,派出坦克纵队穿过布达佩斯,并宣布计划将像格罗夫这样的犹太人运送到工业规模的死亡集中营。几个月后,格罗夫还是个孩子,当红军进军匈牙利首都时,格罗夫再次听到炮声,“解放”了这个国家,强奸了格罗夫的母亲,

Budapest was on the opposite side of the world, but Andy Grove lived through the same typhoon of steel that swept across Asia. Andy (or Andras Grof, as he was then known) survived multiple invasions of Budapest. Hungary’s far-right government treated Jews like the Groves as second-class citizens, but when war broke out in Europe, his father was nevertheless drafted and sent to fight alongside Hungary’s Nazi allies against the Soviet Union, where he was reported missing in action at Stalingrad. Then, in 1944, the Nazis invaded Hungary, their ostensible ally, sending tank columns rolling through Budapest and announcing plans to ship Jews like Grove to industrial-scale death camps. Still a child, Grove heard the thud of artillery again months later as Red Army troops marched into Hungary’s capital, “liberating” the country, raping Grove’s mother, and installing a brutal puppet regime in the Nazis’ place.

无尽的罐柱;飞机波;数千吨炸弹从天而降;运送卡车、战车、石油产品、机车、轨道车、火炮、弹药、煤炭和钢铁的船队——第二次世界大战是一场工业消耗的冲突。美国就是这样想的:一场工业战争是一场美国会赢得的斗争。在华盛顿,随着美国将制造业力量转变为军事力量,战争生产委员会的经济学家们用铜和铁、橡胶和石油、铝和锡来衡量成功。

Endless tank columns; waves of airplanes; thousands of tons of bombs dropped from the skies; convoys of ships delivering trucks, combat vehicles, petroleum products, locomotives, rail cars, artillery, ammunition, coal, and steel—World War II was a conflict of industrial attrition. The United States wanted it that way: an industrial war was a struggle America would win. In Washington, the economists at the War Production Board measured success in terms of copper and iron, rubber and oil, aluminum and tin as America converted manufacturing might into military power.

美国制造的坦克比所有轴心国加起来还多,舰船和飞机数量超过轴心国的两倍,火炮和机枪的产量是轴心国的两倍。工业货物车队从美国港口横渡大西洋和太平洋,为英国、苏联、中国和其他盟国提供关键物资。战争是由斯大林格勒的士兵和中途岛的水手发动的。但战斗力是由美国的凯撒造船厂和红河的装配线产生的。

The United States built more tanks than all the Axis powers combined, more ships, more planes, and twice the Axis production of artillery and machine guns. Convoys of industrial goods streamed from American ports across the Atlantic and Pacific Oceans, supplying Britain, the Soviet Union, China, and other allies with key materiel. The war was waged by soldiers at Stalingrad and sailors at Midway. But the fighting power was produced by America’s Kaiser shipyards and the assembly lines at River Rouge.

1945 年,世界各地的无线电广播宣布战争终于结束。在东京之外,年轻的工程师森田昭夫,穿上全套制服来听裕仁天皇的投降演说,虽然他是一个人听演讲,而不是在其他海军军官的陪伴下,所以他不会被迫承诺仪式自杀。隔着东海,张忠谋庆祝了战争的结束和日本的战败,并迅速回归了与朋友打网球、看电影、玩纸牌的悠闲少年生活。在匈牙利,安迪·格鲁夫和他的母亲慢慢地爬出他们的防空洞,尽管他们在苏联占领期间遭受的痛苦与战争期间一样多。

In 1945, radio broadcasts across the world announced that the war was finally over. Outside of Tokyo, Akio Morita, the young engineer, donned his full uniform to hear Emperor Hirohito’s surrender address, though he listened to the speech alone rather than in the company of other naval officers, so he wouldn’t be pressured to commit ritual suicide. Across the East China Sea, Morris Chang celebrated the war’s end and Japan’s defeat with a prompt return to a leisurely teenaged life of tennis, movies, and card games with friends. In Hungary, Andy Grove and his mother slowly crept out of their bomb shelter, though they suffered as much during the Soviet occupation as during the war itself.

第二次世界大战的结果是由工业产出决定的,但很明显,新技术正在改变军事力量。大国制造了成千上万的飞机和坦克,但他们也建立了研究实验室,开发火箭和雷达等新设备。摧毁广岛和长崎的两颗原子弹引发了很多猜测,即新生的原子时代可能会取代由煤炭和钢铁定义的时代。

World War II’s outcome was determined by industrial output, but it was clear already that new technologies were transforming military power. The great powers had manufactured planes and tanks by the thousands, but they’d also built research labs that developed new devices like rockets and radars. The two atomic bombs that destroyed Hiroshima and Nagasaki brought forth much speculation that a nascent Atomic Age might replace an era defined by coal and steel.

莫里斯·张(Morris Chang)和安迪·格鲁夫(Andy Grove)在 ​​1945 年还是学生,还太小,还没有认真思考过技术或政治。然而,森田昭夫才二十出头,在战争的最后几个月里一直在发展寻热导弹。日本远未部署可行的导弹,但该项目让森田看到了未来。可以设想,战争的胜利不是靠装配线上的铆工,而是靠能够识别目标并自动机动的武器。这个想法看起来像科幻小说,但森田隐约意识到电子计算的新发展可能使机器能够通过解决诸如加法、乘法或求平方根之类的数学问题来“思考”。

Morris Chang and Andy Grove were schoolboys in 1945, too young to have thought seriously about technology or politics. Akio Morita, however, was in his early twenties and had spent the final months of the war developing heat-seeking missiles. Japan was far from fielding workable guided missiles, but the project gave Morita a glimpse of the future. It was becoming possible to envision wars won not by riveters on assembly lines but by weapons that could identify targets and maneuver themselves automatically. The idea seemed like science fiction, but Morita was vaguely aware of new developments in electronic computation that might make it possible for machines to “think” by solving math problems like adding, multiplying, or finding a square root.

当然,使用设备进行计算的想法并不新鲜。自从智人第一次学会数数以来,人们就开始上下翻转手指。古代巴比伦人发明了算盘来操纵大量数字,几个世纪以来,人们通过在这些木网格上来回移动木珠来进行繁衍。期间1800 年代末和 1900 年代初,政府和企业中的大型官僚机构的增长需要军队人类“计算机”,办公室工作人员手持笔、纸,偶尔还有简单的机械计算器——可以加、减、乘、除和计算基本平方根的齿轮箱。

Of course, the idea of using devices to compute wasn’t new. People have flipped their fingers up and down since Homo sapiens first learned to count. The ancient Babylonians invented the abacus to manipulate large numbers, and for centuries people multiplied and divided by moving wooden beads back and forth across these wooden grids. During the late 1800s and early 1900s, the growth of big bureaucracies in government and business required armies of human “computers,” office workers armed with pen, paper, and occasionally simple mechanical calculators—gearboxes that could add, subtract, multiply, divide, and calculate basic square roots.

这些活生生的、会呼吸的计算机可以将工资单制成表格,跟踪销售情况,收集人口普查结果,并筛选为保险单定价所需的火灾和干旱数据。在大萧条时期,美国的工程进展管理局希望雇用失业的上班族,设立了数学表格项目。数百台人类“计算机”坐在曼哈顿办公楼的一排排办公桌前,将对数和指数函数制成表格。该项目出版了 28 卷复杂函数的结果,标题有100,000 到 200,009 的整数倒数表等,共有201 页的数字表。

These living, breathing computers could tabulate payrolls, track sales, collect census results, and sift through the data on fires and droughts that were needed to price insurance policies. During the Great Depression, America’s Works Progress Administration, looking to employ jobless office workers, set up the Mathematical Tables Project. Several hundred human “computers” sat at rows of desks in a Manhattan office building and tabulated logarithms and exponential functions. The project published twenty-eight volumes of the results of complex functions, with titles such as Tables of Reciprocals of the Integers from 100,000 Through 200,009, presenting 201 pages covered in tables of numbers.

有组织的人类计算器组显示了计算的前景,但也显示了使用大脑进行计算的局限性。即使通过使用机械计算器增强了大脑,人类的工作速度也很慢。想要使用数学表格项目的结果的人必须翻阅二十八卷之一的页面才能找到特定对数或指数的结果。需要的计算越多,需要翻阅的页面就越多。

Organized groups of human calculators showed the promise of computation, but also the limits of using brains to compute. Even when brains were enhanced by using mechanical calculators, humans worked slowly. A person looking to use the results of the Mathematical Tables Project had to flip through the pages of one of the twenty-eight volumes to find the result of a specific logarithm or exponent. The more calculations that were needed, the more pages had to be flipped through.

同时,对计算的需求不断增长。甚至在第二次世界大战之前,资金就流入了生产功能更强大的机械计算机的项目中,但战争加速了对计算能力的追捕。一些国家的空军开发了机械轰炸瞄准器来帮助飞行员击中目标。轰炸机机组人员通过转动旋钮来输入风速和高度,这会移动调节玻璃镜的金属杆。这些旋钮和操纵杆比任何飞行员都更准确地“计算”了高度和角度,在飞机归巢时将视线聚焦在目标上。然而,局限性是显而易见的。这样的炸弹瞄准具只考虑了几个输入并提供了一个输出:何时投下炸弹。在完美的测试条件下,美国的炸弹瞄准器比飞行员的猜测更准确。然而,当部署在德国上空时,只有 20% 的美国炸弹落下距离目标一千英尺以内。决定战争的因素是投下的炸弹数量和发射的炮弹数量,而不是机械计算机上试图引导但通常无法引导它们的旋钮。

Meanwhile, the demand for calculations kept growing. Even before World War II, money was flowing into projects to produce more capable mechanical computers, but the war accelerated the hunt for computing power. Several countries’ air forces developed mechanical bombsights to help aviators hit their targets. Bomber crews entered the wind speed and altitude by turning knobs, which moved metal levers that adjusted glass mirrors. These knobs and levers “computed” altitudes and angles more exactly than any pilot could, focusing the sight as the plane homed in on its target. However, the limitations were obvious. Such bombsights only considered a few inputs and provided a single output: when to drop the bomb. In perfect test conditions, America’s bombsights were more accurate than pilots’ guesswork. When deployed in the skies above Germany, though, only 20 percent of American bombs fell within one thousand feet of their target. The war was decided by the quantity of bombs dropped and artillery shells fired, not by the knobs on the mechanical computers that tried and usually failed to guide them.

更高的准确性需要更多的计算。工程师最终开始用电荷取代早期计算机中的机械齿轮。早期的电子计算机使用真空管,一种封装在玻璃中的灯泡状金属灯丝。流过管子的电流可以打开和关闭,其功能类似于算盘珠在木棒上来回移动。打开的电子管被编码为 1,而关闭的真空管被编码为 0。这两个数字可以使用二进制计数系统产生任何数字,因此理论上可以执行多种类型的计算。

More accuracy required more calculations. Engineers eventually began replacing mechanical gears in early computers with electrical charges. Early electric computers used the vacuum tube, a lightbulb-like metal filament enclosed in glass. The electric current running through the tube could be switched on and off, performing a function not unlike an abacus bead moving back and forth across a wooden rod. A tube turned on was coded as a 1 while a vacuum tube turned off was a 0. These two digits could produce any number using a system of binary counting—and therefore could theoretically execute many types of computation.

此外,真空管使这些数字计算机可以重新编程。诸如炸弹瞄准具中的机械齿轮只能执行一种类型的计算,因为每个旋钮都物理地连接到杠杆和齿轮上。算盘上的珠子受到它们来回移动的杆的约束。然而,真空管之间的连接可以重新组织,使计算机能够运行不同的计算。

Moreover, vacuum tubes made it possible for these digital computers to be reprogrammed. Mechanical gears such as those in a bombsight could only perform a single type of calculation because each knob was physically attached to levers and gears. The beads on an abacus were constrained by the rods on which they moved back and forth. However, the connections between vacuum tubes could be reorganized, enabling the computer to run different calculations.

这是计算领域的一次飞跃——如果不是因为飞蛾,它本来就是这样。因为真空管像灯泡一样发光,它们吸引了昆虫,需要定期他们的工程师“调试”。和灯泡一样,真空管也经常烧坏。1945 年,宾夕法尼亚大学为美国陆军建造了一台名为 ENIAC 的最先进的计算机,用于计算火炮弹道,它有 18000 个真空管。平均而言,每两天就有一根管子发生故障,导致整台机器停机,技术人员争先恐后地寻找并更换损坏的部件。ENIAC 可以成倍增加每秒数百个数字,比任何数学家都快。然而它占据了整个房间,因为它的每一个一万八千管是拳头大小。显然,真空管技术太笨重、太慢、太不可靠。只要计算机是飞蛾扑火的怪物,它们就只能用于诸如密码破解之类的小众应用,除非科学家们能找到更小、更快、更便宜的开关。

This was a leap forward in computing—or it would have been, if not for the moths. Because vacuum tubes glowed like lightbulbs, they attracted insects, requiring regular “debugging” by their engineers. Also like lightbulbs, vacuum tubes often burned out. A state-of-the-art computer called ENIAC, built for the U.S. Army at the University of Pennsylvania in 1945 to calculate artillery trajectories, had eighteen thousand vacuum tubes. On average, one tube malfunctioned every two days, bringing the entire machine to a halt and sending technicians scrambling to find and replace the broken part. ENIAC could multiply hundreds of numbers per second, faster than any mathematician. Yet it took up an entire room because each of its eighteen thousand tubes was the size of a fist. Clearly, vacuum tube technology was too cumbersome, too slow, and too unreliable. So long as computers were moth-ridden monstrosities, they’d only be useful for niche applications like code breaking, unless scientists could find a smaller, faster, cheaper switch.

第2章 开关

CHAPTER 2 The Switch

威廉·肖克利(William Shockley)长期以来一直认为,如果要找到更好的“开关”,那就是借助一种称为半导体的材料。肖克利出生在伦敦,是一位环球旅行的采矿工程师,他在加州帕洛阿尔托小镇的果树中长大。作为独生子,他完全相信自己比周围的任何人都优越——他让每个人都知道这一点。他在南加州的加州理工学院上大学,之后在麻省理工学院获得物理学博士学位,并开始在新泽西州的贝尔实验室工作,该实验室当时是世界领先的科学和工程中心之一。他所有的同事都觉得肖克利令人讨厌,但他们也承认他是一位出色的理论物理学家。他的直觉非常准确,以至于肖克利的一位同事说,好像他实际上可以看到电子穿过金属或键合在一起的原子。

William Shockley had long assumed that if a better “switch” was to be found, it would be with the help of a type of material called semiconductors. Shockley, who’d been born in London to a globe-trotting mining engineer, had grown up amid the fruit trees of the sleepy California town of Palo Alto. An only child, he was utterly convinced of his superiority over anyone around him—and he let everyone know it. He went to college at Caltech, in Southern California, before completing a PhD in physics at MIT and starting work at Bell Labs in New Jersey, which at the time was one of the world’s leading centers of science and engineering. All his colleagues found Shockley obnoxious, but they also admitted he was a brilliant theoretical physicist. His intuition was so accurate that one of Shockley’s coworkers said it was as if he could actually see electrons as they zipped across metals or bonded atoms together.

肖克利的专业领域是半导体,是一类独特的材料。大多数材料要么让电流自由流动(如铜线),要么阻止电流(如玻璃)。半导体不同。就其本身而言,硅和锗等半导体材料就像玻璃一样,几乎不导电。但是当添加某些材料并施加电场时,电流可以开始流动。例如,在硅或锗等半导体材料中添加磷或锑会产生负电流。

Semiconductors, Shockley’s area of specialization, are a unique class of materials. Most materials either let electric current flow freely (like copper wires) or block current (like glass). Semiconductors are different. On their own, semiconductor materials like silicon and germanium are like glass, conducting hardly any electricity at all. But when certain materials are added and an electric field is applied, current can begin to flow. Adding phosphorous or antimony to semiconducting materials like silicon or germanium, for example, lets a negative current flow.

用其他元素“掺杂”半导体材料为可以产生和控制电流的新型设备提供了机会。然而,只要它们的电气特性仍然神秘且无法解释,掌握电子在硅或锗等半导体材料中的流动是一个遥远的梦想。直到 1940 年代后期,尽管贝尔实验室积累了所有物理脑力,但没有人能解释为什么半导体材料板会以如此令人费解的方式起作用。

“Doping” semiconductor materials with other elements presented an opportunity for new types of devices that could create and control electric currents. However, mastering the flow of electrons across semiconductor materials like silicon or germanium was a distant dream so long as their electrical properties remained mysterious and unexplained. Until the late 1940s, despite all the physics brainpower accumulated at Bell Labs, no one could explain why slabs of semiconductor materials acted in such puzzling ways.

1945 年,肖克利首先将他所谓的“固态阀”,在他的笔记本上勾勒出一块连接在 90 伏电池上的硅片。他假设在存在电场的情况下放置一块像硅这样的半导体材料可以吸引存储在内部的“自由电子”聚集在半导体边缘附近。如果电场吸引了足够多的电子,半导体的边缘就会变成一种导电材料,就像金属一样,它总是有大量的自由电子。如果是这样,电流可能会开始流过以前根本不导电的材料。肖克利很快就制造了这样一个设备,期望在硅片顶部施加和移除电场可以使其像阀门一样发挥作用,打开和关闭穿过硅的电子流。然而,当他进行这个实验时,他无法检测到结果。“没有什么可衡量的,”他解释道。“很神秘。” 事实上,1940 年代的简单仪器太不精确,无法测量流动的微小电流。

In 1945, Shockley first theorized what he called a “solid state valve,” sketching in his notebook a piece of silicon attached to a ninety-volt battery. He hypothesized that placing a piece of semiconductor material like silicon in the presence of an electric field could attract “free electrons” stored inside to cluster near the edge of the semiconductor. If enough electrons were attracted by the electric field, the edge of the semiconductor would be transformed into a conductive material, like a metal, which always has large numbers of free electrons. If so, an electric current could begin flowing through a material that previously conducted no electricity at all. Shockley soon built such a device, expecting that applying and removing an electric field on top of the piece of silicon could make it function like a valve, opening and closing the flow of electrons across the silicon. When he ran this experiment, however, he was unable to detect a result. “Nothing measurable,” he explained. “Quite mysterious.” In fact, the simple instruments of the 1940s were too imprecise to measure the tiny current that was flowing.

两年后,肖克利在贝尔实验室的两位同事在不同类型的设备上设计了一个类似的实验。肖克利既自豪又令人讨厌的地方,他的同事沃尔特·布拉顿(Walter Brattain)是一位来自华盛顿农村牧场的杰出实验物理学家,约翰·巴丁(John Bardeen)是普林斯顿大学训练的科学家,后来成为唯一获得两项诺贝尔物理学奖的人。谦虚温和。受到肖克利理论的启发,布拉顿和巴丁制造了一个装置,将两根金丝分别通过电线连接到电源和一块金属上,再连接到一块锗上,每根细丝与锗接触的距离不到一毫米。1947 年 12 月 16 日下午,在贝尔实验室的总部,巴丁和布拉顿打开了电源,并能够控制电流风起云涌的锗。肖克利关于半导体材料的理论已被证明是正确的。

Two years later, two of Shockley’s colleagues at Bell Labs devised a similar experiment on a different type of device. Where Shockley was proud and obnoxious, his colleagues Walter Brattain, a brilliant experimental physicist from a cattle ranch in rural Washington, and John Bardeen, a Princeton-trained scientist who’d later become the only person to win two Nobel Prizes in physics, were modest and mild-mannered. Inspired by Shockley’s theorizing, Brattain and Bardeen built a device that applied two gold filaments, each attached by wires to a power source and to a piece of metal, to a block of germanium, with each filament touching the germanium less than a millimeter apart from the other. On the afternoon of December 16, 1947, at Bell Labs’ headquarters, Bardeen and Brattain switched on the power and were able to control the current surging across the germanium. Shockley’s theories about semiconductor materials had been proven correct.

拥有贝尔实验室的 AT&T 从事电话业务,而不是计算机业务,并认为这种设备——很快被命名为“晶体管”——主要是因为它能够放大通过其庞大网络传输电话呼叫的信号。由于晶体管可以放大电流,人们很快意识到,它们将在助听器和收音机等设备中发挥作用,取代可靠性较差的真空管,真空管也用于信号放大。贝尔实验室很快就开始为这种新设备安排专利申请。

AT&T, which owned Bell Labs, was in the business of telephones, not computers, and saw this device—soon christened a “transistor”—as useful primarily for its ability to amplify signals that transmitted phone calls across its vast network. Because transistors could amplify currents, it was soon realized, they would be useful in devices such as hearing aids and radios, replacing less reliable vacuum tubes, which were also used for signal amplification. Bell Labs soon began arranging patent applications for this new device.

肖克利对他的同事们发现了一个实验来证明他的理论感到愤怒,他致力于超越他们。圣诞节期间,他将自己锁在芝加哥的一家酒店房间里两周,并根据他对半导体物理学无与伦比的理解,开始想象不同的晶体管结构。到 1948 年 1 月,他构思了一种新型晶体管,由三块半导体材料组成。外面的两块会有多余的电子;夹在他们之间的那一块会有赤字。如果在夹层的中间层施加一个微小的电流,它会设置一个更大的电流流过整个设备。这种将小电流转换成大电流的过程与 Brattain 和 Bardeen 的晶体管所展示的放大过程相同。但肖克利开始意识到其他用途,类似于他之前理论化的“固态阀”。他可以通过操纵施加在这个晶体管三明治中间的小电流来打开和关闭较大的电流。开关。开关。肖克利设计了一个开关。

Shockley was furious that his colleagues had discovered an experiment to prove his theories, and he was committed to outdoing them. He locked himself in a Chicago hotel room for two weeks over Christmas and began imagining different transistor structures, based on his unparalleled understanding of semiconductor physics. By January 1948, he’d conceptualized a new type of transistor, made up of three chunks of semiconductor material. The outer two chunks would have a surplus of electrons; the piece sandwiched between them would have a deficit. If a tiny current was applied to the middle layer in the sandwich, it set a much larger current flowing across the entire device. This conversion of a small current into a large one was the same amplification process that Brattain and Bardeen’s transistor had demonstrated. But Shockley began to perceive other uses, along the lines of the “solid state valve” he’d previously theorized. He could turn the larger current on and off by manipulating the small current applied to the middle of this transistor sandwich. On, off. On, off. Shockley had designed a switch.

当贝尔实验室在 1948 年 6 月召开新闻发布会宣布其科学家发明了晶体管时,这并不容易理解为什么这些连线的锗块值得特别宣布。《纽约时报》将这个故事埋在第 46 页。《时代》杂志做得更好,以“小脑细胞”为标题报道了这项发明。然而,即使是从不低估自己重要性的肖克利,也无法想象很快会有成千上万、数百万和数十亿个这样的晶体管被​​用于微观尺度在计算任务中取代人脑。

When Bell Labs held a press conference in June 1948 to announce that its scientists had invented the transistor, it wasn’t easy to understand why these wired blocks of germanium merited a special announcement. The New York Times buried the story on page 46. Time magazine did better, reporting the invention under the headline “Little Brain Cell.” Yet even Shockley, who never underestimated his own importance, couldn’t have imagined that soon thousands, millions, and billions of these transistors would be employed at microscopic scale to replace human brains in the task of computing.

第3章 诺伊斯、基尔比和集成电路

CHAPTER 3 Noyce, Kilby, and the Integrated Circuit

晶体管只有在可以简化并大规模销售的情况下才能取代真空管。理论化和发明晶体管只是第一步。现在,挑战是制造成千上万的它们。Brattain 和 Bardeen 对商业或大规模生产几乎没有兴趣。他们本质上是研究人员,在获得诺贝尔奖后,他们继续从事教学和实验。相比之下,肖克利的雄心壮志只增不减。他不仅想出名,还想发财。他告诉朋友,他梦想不仅在《物理评论》等学术刊物上看到自己的名字,而且华尔街日报上,也是。1955 年,他在加利福尼亚州山景城的旧金山郊区建立了 Shockley Semiconductor,就在帕洛阿尔托的街道上,他年迈的母亲仍然住在那里。

The transistor could only replace vacuum tubes if it could be simplified and sold at scale. Theorizing and inventing transistors was simply the first step; now, the challenge was to manufacture them by the thousands. Brattain and Bardeen had little interest in business or mass production. They were researchers at heart, and after winning the Nobel, they continued their careers teaching and experimenting. Shockley’s ambitions, by contrast, only grew. He wanted not only to be famous but also to be rich. He told friends he dreamed of seeing his name not only in academic publications like the Physical Review but in the Wall Street Journal, too. In 1955, he established Shockley Semiconductor in the San Francisco suburb of Mountain View, California, just down the street from Palo Alto, where his aging mother still lived.

肖克利计划制造世界上最好的晶体管,这可能是因为贝尔实验室和晶体管专利的所有者 AT&T 提出将该设备许可给其他公司25,000 美元,对于最尖端的电子技术而言,物超所值。肖克利认为晶体管会有市场,至少可以替代现有电子产品中的真空管。晶体管的电位大小不过,市场尚不清楚。每个人都同意晶体管是基于最先进物理学的一项聪明的技术,但晶体管只有在比真空管做得更好或生产成本更低的情况下才会起飞。肖克利很快将因其关于半导体的理论而获得诺贝尔奖,但如何使晶体管实用和有用的问题是一个工程难题,而不是理论物理学问题。

Shockley planned to build the world’s best transistors, which was possible because AT&T, the owner of Bell Labs and of the transistor patent, offered to license the device to other companies for $25,000, a bargain for the most cutting-edge electronics technology. Shockley assumed that there’d be a market for transistors, at least for replacing vacuum tubes in existing electronics. The potential size of the transistor market, though, was unclear. Everyone agreed transistors were a clever piece of technology based on the most advanced physics, but transistors would take off only if they did something better than vacuum tubes or could be produced more cheaply. Shockley would soon win the Nobel Prize for his theorizing about semiconductors, but the question of how to make transistors practical and useful was an engineering dilemma, not a matter of theoretical physics.

晶体管很快就开始被用来代替计算机中的真空管,但是成千上万个晶体管之间的布线创造了一个复杂的丛林。德州仪器 (TI) 的工程师杰克·基尔比 (Jack Kilby) 于 1958 年夏天在他的德州实验室度过,专注于寻找一种方法来简化由晶体管系统所需的所有电线造成的复杂性。基尔比说话轻声细语、合议、好奇,而且非常聪明。“他从不苛求,”一位同事回忆道。“你知道他想要发生什么,你尽了最大的努力去实现它。” 另一位喜欢与 Kilby 定期共进烧烤午餐的同事说,他是“一个你想见到的最可爱的人”。

Transistors soon began to be used in place of vacuum tubes in computers, but the wiring between thousands of transistors created a jungle of complexity. Jack Kilby, an engineer at Texas Instruments, spent the summer of 1958 in his Texas lab fixated on finding a way to simplify the complexity created by all the wires that systems with transistors required. Kilby was soft-spoken, collegial, curious, and quietly brilliant. “He was never demanding,” one colleague remembered. “You knew what he wanted to have happen and you tried your darndest to make it happen.” Another colleague, who relished regular barbecue lunches with Kilby, said he was “as sweet a guy as you’d ever want to meet.”

Kilby 是贝尔实验室以外最早使用晶体管的人之一,此前他的第一个雇主是位于密尔沃基的 Centralab,从 AT&T 获得技术许可。1958 年,Kilby 离开 Centralab,在德州仪器 (TI) 的晶体管部门工作。TI 总部位于达拉斯,其成立的目的是利用地震波生产设备,以帮助石油工人决定在哪里钻探。二战期间,该公司被美国海军征召制造声纳装置跟踪敌方潜艇。战后,TI 高管意识到这种电子专业知识也可用于其他军事系统,因此他们聘请了像 Kilby 这样的工程师来建造它们。

Kilby was one of the first people outside Bell Labs to use a transistor, after his first employer, Milwaukee-based Centralab, licensed the technology from AT&T. In 1958, Kilby left Centralab to work in the transistor unit of Texas Instruments. Based in Dallas, TI had been founded to produce equipment using seismic waves to help oilmen decide where to drill. During World War II, the company had been drafted by the U.S. Navy to build sonar devices to track enemy submarines. After the war, TI executives realized this electronics expertise could be useful in other military systems, too, so they hired engineers like Kilby to build them.

Kilby 在公司 7 月假期前后抵达达拉斯,但他没有积累假期,所以他独自在实验室里呆了几个星期。随着时间的推移,他想知道如何减少将不同晶体管串在一起所需的电线数量。他没有使用单独的硅或锗来构建每个晶体管,而是考虑组装多个组件在同一块半导体材料上。当他的同事们从暑假回来时,他们意识到 Kilby 的想法是革命性的。多个晶体管可以内置在一块硅或锗中。基尔比称他的发明为“集成电路”,但它被通俗地称为“芯片”,因为每个集成电路都是由一块从圆形硅晶片上“切割”下来的硅制成的。

Kilby arrived in Dallas around the company’s July holiday period, yet he’d accumulated no vacation time so he was left alone in the lab for a couple of weeks. With time to tinker, he wondered how to reduce the number of wires that were needed to string different transistors together. Rather than use a separate piece of silicon or germanium to build each transistor, he thought of assembling multiple components on the same piece of semiconductor material. When his colleagues returned from summer vacation, they realized that Kilby’s idea was revolutionary. Multiple transistors could be built into a single slab of silicon or germanium. Kilby called his invention an “integrated circuit,” but it became known colloquially as a “chip,” because each integrated circuit was made from a piece of silicon “chipped” off a circular silicon wafer.

大约一年前,在加利福尼亚州帕洛阿尔托,威廉·肖克利的半导体实验室雇用的八名工程师告诉他们的诺贝尔奖获得者老板他们要辞职。肖克利有发现人才的本事,但他是一个糟糕的经理。他在争议中茁壮成长,并创造了一种有毒的气氛,疏远了他召集的聪明的年轻工程师。于是这八位工程师离开了肖克利半导体,决定成立自己的公司,Fairchild Semiconductor,获得东海岸百万富翁的种子资金。

About a year earlier, in Palo Alto, California, a group of eight engineers employed by William Shockley’s semiconductor lab had told their Nobel Prize−winning boss that they were quitting. Shockley had a knack for spotting talent, but he was an awful manager. He thrived on controversy and created a toxic atmosphere that alienated the bright young engineers he’d assembled. So these eight engineers left Shockley Semiconductor and decided to found their own company, Fairchild Semiconductor, with seed funding from an East Coast millionaire.

来自肖克利实验室的八名叛逃者被广泛认为是创立硅谷的人。八人之一的 Eugene Kleiner 将继续创立 Kleiner Perkins,这是世界上最强大的风险投资公司之一。继续运行 Fairchild 研发过程的戈登摩尔后来创造了摩尔定律的概念来描述计算能力的指数增长。最重要的是鲍勃·诺伊斯(Bob Noyce),他是“叛徒八人”的领导者,他对微电子具有超凡魅力和远见卓识的热情,并且对需要哪些技术进步才能使晶体管变得小巧、便宜和可靠有一种直觉。将新发明与商业机会相匹配正是像仙童这样的初创公司成功所需要的——也是芯片行业起飞所需要的。

The eight defectors from Shockley’s lab are widely credited with founding Silicon Valley. One of the eight, Eugene Kleiner, would go on to found Kleiner Perkins, one of the world’s most powerful venture capital firms. Gordon Moore, who went on to run Fairchild’s R&D process, would later coin the concept of Moore’s Law to describe the exponential growth in computing power. Most important was Bob Noyce, the leader of the “traitorous eight,” who had a charismatic, visionary enthusiasm for microelectronics and an intuitive sense of which technical advances were needed to make transistors tiny, cheap, and reliable. Matching new inventions with commercial opportunities was exactly what a startup like Fairchild needed to succeed—and what the chip industry needed to take off.

到 Fairchild 成立时,晶体管的科学已经大体清晰,但可靠地制造它们是一项非凡的挑战。第一个商业化的晶体管是由一块锗制成,上面有不同的材料,上面有来自亚利桑那沙漠​​的台地形状。这些层是通过用一滴黑蜡覆盖一部分锗制成的,使用化学蚀刻掉没有被蜡覆盖的锗,然后去除蜡,在锗上形成台面形状。

By the time Fairchild was founded, the science of transistors was broadly clear, but manufacturing them reliably was an extraordinary challenge. The first commercialized transistors were made of a block of germanium with different materials layered on top in the shape of a mesa from the Arizona desert. These layers were fabricated by covering a portion of the germanium with a drop of black wax, using a chemical to etch off the germanium that wasn’t covered with wax, and then removing the wax, creating mesa shapes atop the germanium.

台面结构的一个缺点是它允许灰尘或其他颗粒等杂质滞留在晶体管上,与晶体管表面的材料发生反应。诺伊斯的同事让·霍尔尼(Jean Hoerni)是一位瑞士物理学家和狂热的登山者,他意识到如果整个晶体管可以内置在锗中而不是在锗之上,那么台面就没有必要了。他设计了一种制造晶体管所有部件的方法,方法是在一块硅板上沉积一层保护性二氧化硅,然后在需要的地方蚀刻孔并沉积额外的材料。这种沉积保护层的方法避免了材料暴露在空气和可能导致缺陷的杂质中。这是可靠性方面的重大进步。

A downside of the mesa structure was that it allowed impurities like dust or other particles to become lodged on the transistor, reacting with the materials on the its surface. Noyce’s colleague Jean Hoerni, a Swiss physicist and avid mountaineer, realized the mesas weren’t necessary if the entire transistor could be built into, rather than on top of, the germanium. He devised a method of fabricating all the parts of a transistor by depositing a layer of protective silicon dioxide on top of a slab of silicon, then etching holes where needed and depositing additional materials. This method of depositing protective layers avoided exposing materials to air and impurities that could cause defects. It was a major advance in reliability.

几个月后,Noyce 意识到 Hoerni 的“平面方法”可用于在同一块硅片上生产多个晶体管。诺伊斯不知道,基尔比在锗基上制造了一个台面晶体管,然后用电线将其连接起来,诺伊斯使用 Hoerni 的平面工艺在同一芯片上构建多个晶体管。由于平面工艺用二氧化硅绝缘层覆盖晶体管,诺伊斯可以通过在其顶部沉积金属线将“导线”直接放置在芯片上,从而在芯片的晶体管之间导电。和 Kilby 一样,Noyce 也生产了一种集成电路:在一块半导体材料上拥有多个电子元件。然而,诺伊斯的版本根本没有独立的电线。晶体管被内置在一块材料中。很快,

Several months later, Noyce realized Hoerni’s “planar method” could be used to produce multiple transistors on the same piece of silicon. Where Kilby, unbeknownst to Noyce, had produced a mesa transistor on a germanium base and then connected it with wires, Noyce used Hoerni’s planar process to build multiple transistors on the same chip. Because the planar process covered the transistor with an insulating layer of silicon dioxide, Noyce could put “wires” directly on the chip by depositing lines of metal on top of it, conducting electricity between the chip’s transistors. Like Kilby, Noyce had produced an integrated circuit: multiple electric components on a single piece of semiconductor material. However, Noyce’s version had no freestanding wires at all. The transistors were built into a single block of material. Soon, the “integrated circuits” that Kilby and Noyce had developed would become known as “semiconductors” or, more simply, “chips.”

诺伊斯、摩尔和他们在飞兆半导体的同事知道他们的集成电路将比其他电子设备所依赖的迷宫般的电线可靠得多。将飞兆半导体的“平面”设计小型化似乎比标准台面晶体管要容易得多。与此同时,较小的电路将需要更少的电力来工作。诺伊斯和摩尔开始意识到小型化和电效率是一个强大的组合:更小的晶体管和更低的功耗将为它们的集成电路创造新的用例。然而,一开始,诺伊斯的集成电路成本与将单独的组件连接在一起的简单设备相比,其制造量要高出 50 倍。每个人都同意诺伊斯的发明很聪明,甚至很出色。它所需要的只是一个市场。

Noyce, Moore, and their colleagues at Fairchild Semiconductor knew their integrated circuits would be vastly more reliable than the maze of wires that other electronic devices relied on. It seemed far easier to miniaturize Fairchild’s “planar” design than standard mesa transistors. Smaller circuits, meanwhile, would require less electricity to work. Noyce and Moore began to realize that miniaturization and electric efficiency were a powerful combination: smaller transistors and reduced power consumption would create new use cases for their integrated circuits. At the outset, however, Noyce’s integrated circuit cost fifty times as much to make as a simpler device with separate components wired together. Everyone agreed Noyce’s invention was clever, even brilliant. All it needed was a market.

第 4 章 起飞

CHAPTER 4 Liftoff

在诺伊斯和摩尔创立仙童半导体三天后的晚上 8 点 55 分,谁来为集成电路买单的问题的答案在加利福尼亚的夜空中飞驰而过。人造卫星是世界上第一颗人造卫星,由苏联发射,以每小时一万八千英里的速度从西向东绕地球运行。《旧金山纪事报》的标题是“俄罗斯'月球'环绕地球” ,这反映了美国人担心这颗卫星会给俄罗斯人带来战略优势。四年后,当宇航员尤里·加加林成为太空第一人时,苏联再次震惊了人造卫星。

Three days after Noyce and Moore founded Fairchild Semiconductor, at 8:55 p.m., the answer to the question of who would pay for integrated circuits hurtled over their heads through California’s nighttime sky. Sputnik, the world’s first satellite, launched by the Soviet Union, orbited the earth from west to east at a speed of eighteen thousand miles per hour. “Russ ‘Moon’ Circling Globe,” declared the headline in the San Francisco Chronicle, reflecting Americans’ fears that this satellite gave the Russians a strategic advantage. Four years later, the Soviet Union followed Sputnik with another shock when cosmonaut Yuri Gagarin became the first person in space.

在美国各地,苏联的太空计划引起了信任危机。控制宇宙会产生严重的军事后果。美国以为自己是世界科学超级大国,现在看来已经落伍了。华盛顿启动了一项紧急计划以赶上苏联的火箭和导弹计划,约翰·肯尼迪总统宣布美国将派人登月。Bob Noyce 的集成电路突然有了一个市场:火箭。

Across America, the Soviet space program caused a crisis of confidence. Control of the cosmos would have serious military ramifications. The U.S. thought it was the world’s science superpower, but now it seemed to have fallen behind. Washington launched a crash program to catch up with the Soviets’ rocket and missile programs, and President John F. Kennedy declared the U.S. would send a man to the moon. Bob Noyce suddenly had a market for his integrated circuits: rockets.

Noyce 芯片的第一笔大订单来自 NASA,它在 1960 年代有巨额预算将宇航员送上月球。当美国将目光投向登月时,美国宇航局委托麻省理工学院仪器实验室的工程师设计制导计算机阿波罗宇宙飞船,这个设备肯定是有史以来最复杂的计算机之一。每个人都同意基于晶体管的计算机比二战期间破解密码和计算炮弹轨迹的真空管计算机要好得多。但是这些设备中的任何一个真的可以引导航天器前往月球吗?麻省理工学院的一位工程师计算出,为了满足阿波罗任务的需要,一台计算机需要有冰箱那么大,并且消耗的电力比整个阿波罗宇宙飞船预计产生的电力还要多。

The first big order for Noyce’s chips came from NASA, which in the 1960s had a vast budget to send astronauts to the moon. As America set its sights on a lunar landing, engineers at the MIT Instrumentation Lab were tasked by NASA to design the guidance computer for the Apollo spacecraft, a device that was certain to be one of the most complicated computers ever made. Everyone agreed transistor-based computers were far better than the vacuum-tube equivalents that had cracked codes and calculated artillery trajectories during World War II. But could any of these devices really guide a spacecraft to the moon? One MIT engineer calculated that to meet the needs of the Apollo mission, a computer would need to be the size of a refrigerator and would consume more electricity than the entire Apollo spacecraft was expected to produce.

麻省理工学院的仪器实验室在 1959 年收到了由德州仪器公司生产的第一块集成电路,就在杰克·基尔比发明它的一年后,他以 1,000 美元的价格购买了 64 个这样的芯片,作为美国海军导弹的一部分进行测试程序。麻省理工学院的团队最终没有在该导弹中使用芯片,但发现集成电路的想法很有趣。大约在同一时间,Fairchild 开始销售自己的“Micrologic”芯片。1962 年 1 月,麻省理工学院的一位工程师命令一位同事:“出去大量购买这些东西,”看看它们是不是真的。”

MIT’s Instrumentation Lab had received its first integrated circuit, produced by Texas Instruments, in 1959, just a year after Jack Kilby had invented it, buying sixty-four of these chips for a price of $1,000 to test them as part of a U.S. Navy missile program. The MIT team ended up not using chips in that missile but found the idea of integrated circuits intriguing. Around the same time, Fairchild began marketing its own “Micrologic” chips. “Go out and buy large quantities of those things,” one MIT engineer ordered a colleague in January 1962, “to see if they are real.”

Fairchild 是一家全新的公司,由一群 30 岁的工程师经营,没有任何业绩记录,但他们的芯片可靠且准时到达。到 1962 年 11 月,管理 MIT 实验室的著名工程师查尔斯·斯塔克·德雷珀(Charles Stark Draper)决定为阿波罗计划押注仙童芯片,计算得出使用诺伊斯集成电路的计算机将比基于分立晶体管的计算机小三分之一且更轻。它也将使用更少的电力。最终将阿波罗 11 号送上月球的计算机重达 70 磅,占用约 1 立方英尺的空间,比二战期间计算炮弹轨迹的宾夕法尼亚大学 ENIAC 计算机少一千倍。

Fairchild was a brand-new company, run by a group of thirty-year-old engineers with no track record, but their chips were reliable and arrived on time. By November 1962, Charles Stark Draper, the famed engineer who ran the MIT lab, had decided to bet on Fairchild chips for the Apollo program, calculating that a computer using Noyce’s integrated circuits would be one-third smaller and lighter than a computer based on discrete transistors. It would use less electricity, too. The computer that eventually took Apollo 11 to the moon weighed seventy pounds and took up about one cubic foot of space, a thousand times less than the University of Pennsylvania’s ENIAC computer that had calculated artillery trajectories during World War II.

麻省理工学院认为阿波罗制导计算机是其最自豪的成就之一,但鲍勃·诺伊斯知道是他的芯片让阿波罗计算机运转起来。诺伊斯吹嘘说,到 1964 年,阿波罗计算机中的集成电路已经运行了 1900 万小时,而只有两个故障,其中一个是由移动计算机时的物理损坏引起的。阿波罗计划的芯片销售将飞兆半导体从一家小型初创公司转变为拥有一千名员工的公司。销售额从 1958 年的 500,000 美元激增至两年后达到 2100 万美元。

MIT considered the Apollo guidance computer one of its proudest accomplishments, but Bob Noyce knew that it was his chips that made the Apollo computer tick. By 1964, Noyce bragged, the integrated circuits in Apollo computers had run for 19 million hours with only two failures, one of which was caused by physical damage when a computer was being moved. Chip sales to the Apollo program transformed Fairchild from a small startup into a firm with one thousand employees. Sales ballooned from $500,000 in 1958 to $21 million two years later.

随着诺伊斯为 NASA 增加产量,他降低了其他客户的价格。1961 年 12 月售价 120 美元的集成电路是到明年 10 月打折到 15 美元。NASA 对引导宇航员登月的集成电路的信任是认可的重要标志。Fairchild 的 Micrologic 芯片不再是未经测试的技术。它们被用于最无情和最崎岖的环境:外太空。

As Noyce ramped up production for NASA, he slashed prices for other customers. An integrated circuit that sold for $120 in December 1961 was discounted to $15 by next October. NASA’s trust in integrated circuits to guide astronauts to the moon was an important stamp of approval. Fairchild’s Micrologic chips were no longer an untested technology; they were used in the most unforgiving and rugged environment: outer space.

这对 Jack Kilby 和德州仪器来说是个好消息,尽管他们的芯片在阿波罗计划中只发挥了很小的作用。在达拉斯的 TI 总部,Kilby 和 TI 总裁 Pat Haggerty 正在为他们自己的集成电路寻找大客户。哈格蒂是一名来自南达科他州小镇的铁路电报员的儿子,他曾接受过电气工程师培训,并在二战期间为美国海军从事电子产品工作。自从 1951 年加入德州仪器以来,哈格蒂一直专注于向军方出售电子系统。

This was good news for Jack Kilby and Texas Instruments, even though their chips played only a small role in the Apollo program. At TI headquarters in Dallas, Kilby and TI president Pat Haggerty were looking for a big customer for their own integrated circuits. Haggerty was the son of a railroad telegrapher from small-town South Dakota who’d trained as an electrical engineer and worked on electronics for the U.S. Navy during World War II. Since the day he arrived at Texas Instruments in 1951, Haggerty had focused on selling electronic systems to the military.

哈格蒂直觉地明白,杰克·基尔比的集成电路最终可以插入美国军方使用的每一件电子产品。一位迷人的公众演说家,当他向德州仪器 (TI) 员工宣讲电子产品的未来时,一位德州仪器 (TI) 资深人士将 Haggerty 铭记为“就像在山顶上说话的救世主。他似乎可以预测一切。” 当美国和苏联在 1960 年代初陷入核僵局时——首先是为了控制分裂的柏林,然后是古巴导弹危机——哈格蒂没有比五角大楼更好的客户了。就在基尔比创建集成电路几个月后,哈格蒂向国防部工作人员简要介绍了基尔比的发明。第二年,空军航空电子实验室同意赞助 TI 的芯片研究。随后有几份小型军事设备合同。但哈格蒂正在寻找一条大鱼。

Haggerty intuitively understood that Jack Kilby’s integrated circuit could eventually be plugged into every piece of electronics the U.S. military used. A captivating public speaker, when he preached to Texas Instruments employees about the future of electronics, Haggerty was remembered by one TI veteran as “like a messiah speaking from the mountaintop. He seemed like he could predict everything.” As the U.S. and the Soviet Union lurched between nuclear standoffs in the early 1960s—first over control of divided Berlin, then during the Cuban Missile Crisis—Haggerty had no better customer than the Pentagon. Just months after Kilby created the integrated circuit, Haggerty briefed Defense Department staff on Kilby’s invention. The next year, the Air Force Avionics Lab agreed to sponsor TI’s chip research. Several small contracts for military devices followed. But Haggerty was looking for a big fish.

1962 年秋天,空军开始寻找一种新的计算机来引导其民兵 II 导弹,该导弹旨在在攻击苏联之前将核弹头投掷穿过太空。民兵的第一个版本刚刚投入使用,但它太重了,几乎无法从散布在美国西部的发射场击中莫斯科。它的机载制导计算机是一个庞大的怪物,基于离散晶体管,目标程序通过以下方式输入制导计算机带有打孔的聚酯薄膜胶带。

In fall 1962, the Air Force began looking for a new computer to guide its Minuteman II missile, which was designed to hurl nuclear warheads through space before striking the Soviet Union. The first version of the Minuteman had just entered service, but it was so heavy it could barely hit Moscow from launch sites scattered across the American West. Its onboard guidance computer was a hulking monstrosity, based on discrete transistors, with the targeting program fed into the guidance computer via Mylar tape with holes punched in it.

哈格蒂向空军承诺,使用 Kilby 集成电路的计算机可以以一半的重量执行两倍的计算。他设想了一台使用 22 种不同类型集成电路的计算机。在他看来,计算机 95% 的功能将由刻在硅中的集成电路来执行,这些集成电路总重 2.2 盎司。剩下的 5% 的计算机硬件重达 36 磅,TI 的工程师还无法弄清楚如何将其安装在芯片上。“这只是尺寸和重量的问题,”一位设计计算机的工程师 Bob Nease 在谈到使用集成电路的决定时解释道。“真的没有太多选择。”

Haggerty promised the Air Force that a computer using Kilby’s integrated circuits could perform twice the computations with half the weight. He envisioned a computer that used twenty-two different types of integrated circuits. In his mind’s eye, 95 percent of the computer’s functions would be conducted by integrated circuits carved into silicon, which together weighed 2.2 ounces. The remaining 5 percent of the computer hardware, which TI’s engineers couldn’t yet figure out how to put on a chip, weighed 36 pounds. “It was just a matter of size and weight,” explained one the engineers designing the computer, Bob Nease, regarding the decision to use integrated circuits. “There was really not much of a choice.”

赢得 Minuteman II 合同改变了 TI 的芯片业务。TI 的集成电路销售量以前是几十个,但由于担心美国与苏联存在“导弹差距”,该公司很快就以数千个销售。一年之内,TI 向空军的出货量占迄今为止购买芯片的所有美元的 60%。到 1964 年底,德州仪器已经为民兵计划提供了十万块集成电路。到 1965 年,当年销售的所有集成电路中有 20% 用于民兵计划。帕特·哈格蒂 (Pat Haggerty) 向军方出售芯片的赌注得到了回报。唯一的问题是 TI 是否可以学习如何批量生产它们。

Winning the Minuteman II contract transformed TI’s chip business. TI’s integrated circuit sales had previously been measured in the dozens, but the firm was soon selling them by the thousands amid fear of an American “missile gap” with the Soviet Union. Within a year, TI’s shipments to the Air Force accounted for 60 percent of all dollars spent buying chips to date. By the end of 1964, Texas Instruments had supplied one hundred thousand integrated circuits to the Minuteman program. By 1965, 20 percent of all integrated circuits sold that year went to the Minuteman program. Pat Haggerty’s bet on selling chips to the military was paying off. The only question was whether TI could learn how to mass-produce them.

第 5 章 迫击炮和批量生产

CHAPTER 5 Mortars and Mass Production

Ĵ1958 年 9 月 1 日,Lathrop 进入德州仪器公司的停车场开始他的第一天工作,恰逢 Jack Kilby 在 TI 实验室修修补补的决定性夏天即将结束。从麻省理工学院毕业后,他与鲍勃·诺伊斯(Bob Noyce)合作,在美国政府实验室工作,他的任务是设计一种近炸引信,使 81 毫米迫击炮弹能够在目标上方自动引爆。和 Fairchild 的工程师一样,他一直在努力解决难以小型化的台面形晶体管。现有的制造工艺包括在半导体材料的某些部分上放置特殊形状的蜡球,然后使用专门的化学品清洗未覆盖的部分。制造更小的晶体管需要更小的蜡球,

Jay Lathrop pulled into Texas Instruments’ parking lot for his first day of work on September 1, 1958, just as Jack Kilby’s fateful summer spent tinkering in TI’s labs was coming to a close. After graduating from MIT, where he’d overlapped with Bob Noyce, Lathrop had worked at a U.S. government lab where he was tasked with devising a proximity fuse that would enable an 81mm mortar shell to detonate automatically above its target. Like engineers at Fairchild, he was struggling with mesa-shaped transistors, which were proving difficult to miniaturize. Existing manufacturing processes involved placing specially shaped globs of wax on certain portions of the semiconductor material, then washing away the uncovered portions using specialized chemicals. Making smaller transistors required smaller globs of wax, but keeping these globs in the correct shape proved challenging.

在通过显微镜观察他们的一个晶体管时,Lathrop 和他的助手、化学家 James Nall 有了一个想法:显微镜镜头可以把一些小东西放大,让它看起来更大。如果他们把显微镜倒过来,它的镜头会占用一些大的东西,让它看起来更小。他们能否使用镜头拍摄大图案并将其“打印”到锗上,从而在锗块上制作微型台面?相机公司柯达出售称为光刻胶的化学物质,这种化学物质在曝光时会发生反应。

While looking through a microscope at one of their transistors, Lathrop and his assistant, chemist James Nall, had an idea: a microscope lens could take something small and make it look bigger. If they turned the microscope upside down, its lens would take something big and make it look smaller. Could they use a lens to take a big pattern and “print” it onto germanium, thereby making miniature mesas on their blocks of germanium? Kodak, the camera company, sold chemicals called photoresists, which reacted when exposed to light.

Lathrop 用柯达的一种光刻胶化学物质覆盖了一块锗,如果暴露在光线下,这种化学物质就会消失。接下来,他将显微镜倒置,用图案覆盖镜头,使光线只能通过矩形区域。光线进入图案,通过透镜以矩形形状照射,当它聚焦在涂有光刻胶的锗上时,通过倒置显微镜缩小尺寸,光线形成完美形状的微型矩形版本图案。在光线照射到光刻胶层的地方,化学结构发生了变化,使其被冲走,留下一个微小的矩形孔,比任何蜡球都小得多,形状也更准确。很快,Lathrop 发现他也可以打印“电线”,

Lathrop covered a block of germanium with one of Kodak’s photoresist chemicals that would disappear if exposed to light. Next, he turned his microscope upside down, covering the lens with a pattern so that light would only pass through a rectangle-shaped area. Light entered the pattern, shined in a rectangle shape through the lens, and was shrunk in size by the upside-down microscope as it focused onto the photoresist-coated germanium, with the rays of light creating a perfectly shaped, miniature version of the rectangular pattern. Where light struck the layer of photoresist, the chemical structure was altered, allowing it to be washed away, leaving a tiny rectangular hole, far smaller and more accurately shaped than any glob of wax could have been. Soon Lathrop discovered he could print “wires,” too, by adding an ultra-thin layer of aluminum to connect the germanium with an external power source.

Lathrop 将这一过程称为光刻——用光打印。他生产的晶体管比以前可能的要小得多,直径只有十分之一英寸,高度只有 0.0005 英寸。光刻技术使大规模生产微型晶体管成为可能。Lathrop 于 1957 年申请了这项技术的专利。在陆军乐队演奏时,军方给了他一枚奖章和 25,000 美元的现金奖金,他用这笔钱给家人买了一辆 Nash Rambler 旅行车。

Lathrop called the process photolithography—printing with light. He produced transistors much smaller than had previously been possible, measuring only a tenth of an inch in diameter, with features as small as 0.0005 inches in height. Photolithography made it possible to imagine mass-producing tiny transistors. Lathrop applied for a patent on the technique in 1957. With the Army band playing, the military gave him a medal for his work and a $25,000 cash bonus, which he used to buy his family a Nash Rambler station wagon.

Pat Haggerty 和 Jack Kilby 立即意识到 Lathrop 的光刻工艺价值远远超过军队给予他的 25,000 美元奖金。Minuteman II 导弹计划需要数千个集成电路。阿波罗飞船还需要数万人。Haggerty 和 Kilby 意识到光线和光刻胶可以解决大规模生产问题,以手工焊接导线无法实现的方式实现芯片制造的机械化和小型化。

Pat Haggerty and Jack Kilby immediately realized Lathrop’s photolithography process was worth a lot more than the $25,000 prize the Army had given him. The Minuteman II missile program needed thousands of integrated circuits. The Apollo spacecraft needed tens of thousands more. Haggerty and Kilby realized that light rays and photoresists could solve the mass-production problem, mechanizing and miniaturizing chipmaking in a way that soldering wires together by hand could not.

在德州仪器实施 Lathrop 的光刻工艺需要新材料和新工艺。柯达的光刻胶化学品纯度不足以进行大规模生产,因此 TI 购买了自己的离心机并重新加工了柯达提供的化学品。拉斯罗普他乘坐火车穿越全国,寻找“掩模”,可以用来将精确的光图案投射到覆盖有光刻胶的半导体材料板上以雕刻电路。他最终得出结论,现有的口罩公司都没有足够的精度,因此 TI 也决定自己制造口罩。Kilby 的集成电路所需的硅片必须是超纯的,超出任何公司的销售量。因此,TI 也开始生产自己的硅晶圆。

Implementing Lathrop’s lithography process at Texas Instruments required new materials and new processes. Kodak’s photoresist chemicals were insufficiently pure for mass production, so TI bought its own centrifuges and reprocessed the chemicals that Kodak supplied. Lathrop took trains across the country in search of “masks” that could be used to project precise patterns of light onto photoresist-covered slabs of semiconductor material to carve circuits. He eventually concluded that no existing mask company had sufficient precision, so TI decided to make masks itself, too. The slabs of silicon that Kilby’s integrated circuits required had to be ultra-pure, beyond what any company sold. TI therefore also began producing its own silicon wafers.

当一切都标准化时,大规模生产工作。通用汽车将许多相同的汽车零件插入所有从其装配线下线的雪佛兰。在半导体方面,像 TI 这样的公司缺乏了解其集成电路的所有组件是否相同的工具。化学品含有当时无法测试的杂质。温度和压力的变化引起了意想不到的化学反应。投射光线的面罩可能被灰尘颗粒污染。一个杂质就可能毁掉整个生产过程。唯一的改进方法是反复试验,TI 组织了数千次实验来评估不同温度、化学组合和生产过程的影响。Jack Kilby 每个星期六都在 TI 的走廊里踱步,检查工程师的实验。

Mass production works when everything is standardized. General Motors plugged many of the same car parts into all the Chevrolets that rolled off its assembly lines. When it came to semiconductors, companies like TI lacked the tools to know whether all the components of their integrated circuits were the same. Chemicals had impurities that at the time were impossible to test. Variation in temperature and pressure caused unexpected chemical reactions. The masks through which light was projected could be contaminated by particles of dust. A single impurity could ruin an entire production run. The only method of improvement was trial and error, with TI organizing thousands of experiments to assess the impact of different temperatures, chemical combinations, and production processes. Jack Kilby spent each Saturday pacing TI’s hallways and checking on his engineers’ experiments.

TI 生产工程师玛丽安妮波特花了几个月的时间进行全天候测试。作为第一位从德克萨斯理工大学获得物理学学位的女性,Potter 受聘于 TI,负责扩大民兵导弹的芯片生产。她经常上夜班,从晚上 11 点到早上 8 点,以确保实验按计划进行。收集数据需要几天的实验。然后她对数据进行回归,使用她的计算尺计算指数和平方根,将结果绘制在图表上,然后解释它们——这一切都是手工完成的。这是一个缓慢、费力、痛苦的过程,依靠人类“计算机”来处理数字。然而,反复试验是德州仪器唯一的方法。

TI production engineer Mary Anne Potter spent months running round-the-clock tests. The first woman to earn a physics degree from Texas Tech, Potter was hired at TI to scale up chip production for the Minuteman missile. She often worked the night shift, from 11 p.m. until 8 a.m., to make sure experiments were progressing according to plan. Gathering data took days of experimentation. Then she ran regressions on the data, using her slide rule to calculate exponents and square roots, plot the results on a graph, and then interpret them—doing it all by hand. It was a slow, laborious, painful process, relying on human “computers” to crunch numbers. Yet trial and error was the only method Texas Instruments had.

Morris Chang 于 1958 年与 Jay Lathrop 同年加入 TI,并负责晶体管生产线。近一个自从张先生逃离上海以逃避前进的共产主义军队以来,已经过去了十年,他先是去了香港,然后是波士顿,他获得了哈佛的录取,他是哈佛大一新生中唯一的中国学生。在学习莎士比亚一年后,张开始担心自己的职业前景。“有美籍华人洗衣店,也有美籍华人餐馆,”他回忆道。“50 年代初期,华裔美国人唯一可以从事的真正严肃的……中产阶级职业是技术。” 机械工程似乎比英语文学更安全,Chang 决定,所以他转到了麻省理工学院。

Morris Chang arrived at TI in 1958, the same year as Jay Lathrop, and was put in charge of a production line of transistors. Nearly a decade had passed since Chang fled Shanghai to escape the advancing Communist armies, first to Hong Kong and then to Boston, having won admission to Harvard, where he was the only Chinese student in the freshman class. After a year spent studying Shakespeare, Chang began to worry about his career prospects. “There were Chinese-American laundry people, there were Chinese-American restaurant people,” he recalled. “The only really serious… middle class profession that a Chinese American could pursue in the early fifties was technical.” Mechanical engineering seemed safer than English literature, Chang decided, so he transferred to MIT.

毕业后,Chang 被 Sylvania 聘用,这是一家在波士顿以外设有工厂的大型电子公司。他的任务是提高 Sylvania 的制造“产量”——实际工作的晶体管的份额。Chang 每天都在修补 Sylvania 的生产工艺,晚上则研究 Shockley 的《半导体中的电子和空穴》,早期半导体电子学的圣经。在 Sylvania 工作了三年后,Chang 收到了 TI 的工作邀请,并搬到了德克萨斯州的达拉斯——他记得是“牛仔之乡”,还有一块“95 美分牛排”的土地。他回忆说,他的任务是运行一条用于 IBM 计算机的晶体管生产线,这种晶体管非常不可靠,以至于 TI 的产量接近于零。几乎所有产品都存在导致电路短路或故障的制造缺陷;他们必须是抛出了。

After graduating, Chang was hired by Sylvania, a big electronics firm with facilities outside of Boston. He was tasked with improving Sylvania’s manufacturing “yield”—the share of transistors that actually worked. Chang spent his days tinkering with Sylvania’s production processes and his evenings studying Shockley’s Electrons and Holes in Semiconductors, the bible of early semiconductor electronics. After three years at Sylvania, Chang received a job offer from TI, and moved to Dallas, Texas—“cowboy country,” he remembered, and a land of “95-cent steaks.” He was tasked with running a production line of transistors to be used in IBM computers, a type of transistor so unreliable that TI’s yield was close to zero, he recalled. Almost all had manufacturing imperfections that caused circuits to short or to malfunction; they had to be tossed out.

作为一名桥牌大师,Chang 像玩他最喜欢的纸牌游戏一样有条不紊地进行制造。到达 TI 后,他开始系统地调整不同化学物质组合时的温度和压力,以确定哪种组合效果最好,并将他的直觉应用到数据中,这种方式让他的同事们感到惊讶和害怕。“当你和他一起工作时,你必须小心,”一位同事回忆道。“他坐在那里,抽着烟斗,隔着烟雾看着你。” 为他工作的德州人认为他“像佛一样”。烟草烟雾的背后是首屈一指的大脑。“他对固态物理学的了解足以驾驭任何人,”一位同事回忆道。他有以强硬的老板而闻名。“莫里斯打人太糟糕了,”一名下属回忆道。“如果你没有被莫里斯骂过,那你就没有去过 TI。” 不过,张的方法产生了结果。几个月内,他的晶体管生产线的产量跃升至 25%。美国最大的科技公司 IBM 的高管来到达拉斯学习他的方法。不久,他被任命负责 TI 的整个集成电路业务。

A master bridge player, Chang approached manufacturing as methodically as he played his favorite card game. Upon arriving at TI, he began systematically tweaking the temperature and pressure at which different chemicals were combined, to determine which combinations worked best, applying his intuition to the data in a way that amazed and intimidated his colleagues. “You had to be careful when you worked with him,” remembered one colleague. “He sat there and puffed on his pipe and looked at you through the smoke.” The Texans who worked for him thought he was “like a Buddha.” Behind the tobacco smoke was a brain second to none. “He knew enough about solid-state physics to lord it over anyone,” one colleague recalled. He had a reputation for being a tough boss. “Morris was so bad for beating up on people,” one subordinate recalled. “If you hadn’t ever been chewed out by Morris, you hadn’t been at TI.” Chang’s methods produced results, though. Within months, the yield on his production line of transistors jumped to 25 percent. Executives from IBM, America’s biggest tech company, came to Dallas to study his methods. Soon he was placed in charge of TI’s entire integrated circuit business.

和 Chang 一样,Noyce 和 Moore 认为芯片行业的发展没有限制,只要他们能够搞定大规模生产。诺伊斯意识到他的麻省理工学院同学杰伊·拉斯罗普(Jay Lathrop)发现了一种可以改变晶体管制造的技术。Noyce 迅速采取行动,聘请 Lathrop 的实验室合作伙伴化学家 James Nall 在 Fairchild 开发光刻技术。“除非我们能够让它发挥作用,”诺伊斯解释说,“否则我们就没有公司。”

Like Chang, Noyce and Moore saw no limits to the growth of the chip industry so long as they could figure out mass production. Noyce realized his MIT classmate Jay Lathrop, with whom he’d hiked New Hampshire’s mountains while in graduate school, had discovered a technique that could transform transistor manufacturing. Noyce acted swiftly to hire Lathrop’s lab partner, chemist James Nall, to develop photolithography at Fairchild. “Unless we could make it work,” Noyce reasoned, “we did not have a company.”

像 Andy Grove 这样的生产工程师需要改进 Fairchild 的制造工艺。1956 年逃离匈牙利共产主义政府并以难民身份抵达纽约后,格鲁夫在伯克利攻读博士学位。他曾在 1962 年写信给 Fairchild 要求面试,但被告知稍后再试:“我们希望我们的年轻人在完成与其他人的面试后再与我们面试,”拒绝信解释道。格罗夫回忆道,格鲁夫发现仙童的拒绝信“令人作呕”,这是定义硅谷的狂妄自大的早期迹象。但随着对仙童半导体的需求增加,该公司突然急需化学工程师。一位公司高管打电话给伯克利,询问化学系最优秀的学生名单。“那是一见钟情,”格鲁夫回忆道。他于 1963 年被聘用,余生将与诺伊斯和摩尔一起建立芯片行业。

It was up to production engineers like Andy Grove to improve Fairchild’s manufacturing process. After fleeing Hungary’s Communist government in 1956 and arriving in New York as a refugee, Grove had worked his way into a PhD program at Berkeley. He’d written Fairchild in 1962 to ask for a job interview but was told to try again later: “We like our young men to interview with us when they have finished interviewing with everybody else,” the rejection letter explained. Grove found Fairchild’s rejection letter “condescendingly disgusting,” he recalled, an early sign of the hubris that would come to define Silicon Valley. But as demand for Fairchild’s semiconductors increased, the company suddenly had a desperate need for chemical engineers. One company executive rang Berkeley and asked for a list of the best students in the Chemistry Department. Grove was at the top of the list and was called to Palo Alto to meet Gordon Moore. “It was love at first sight,” Grove remembered. He was hired in 1963 and would spend the rest of his life building the chip industry alongside Noyce and Moore.

发明晶体管的诺贝尔奖授予了肖克利、巴丁和布拉顿。杰克·基尔比后来因创造了第一个集成电路; 如果 Bob Noyce 没有在 62 岁时去世,他会与 Kilby 分享奖金。这些发明至关重要,但仅靠科学还不足以建立芯片产业。与学术物理学一样,聪明的制造技术使半导体的传播成为可能。麻省理工学院和斯坦福大学等大学在发展半导体知识方面发挥了至关重要的作用,但芯片行业之所以起飞,是因为这些机构的毕业生花了数年时间调整生产流程,以使大规模制造成为可能。正是工程和直觉,以及科学理论,将贝尔实验室的专利变成了一个改变世界的行业。

The Nobel Prize for inventing the transistor went to Shockley, Bardeen, and Brattain. Jack Kilby later won a Nobel for creating the first integrated circuit; had Bob Noyce not died at the age of sixty-two, he’d have shared the prize with Kilby. These inventions were crucial, but science alone wasn’t enough to build the chip industry. The spread of semiconductors was enabled as much by clever manufacturing techniques as academic physics. Universities like MIT and Stanford played a crucial role in developing knowledge about semiconductors, but the chip industry only took off because graduates of these institutions spent years tweaking production processes to make mass manufacturing possible. It was engineering and intuition, as much as scientific theorizing, that turned a Bell Labs patent into a world-changing industry.

肖克利被公认为他这一代最伟大的理论物理学家之一,最终放弃了发家致富的努力,并在《华尔街日报》上名声大噪。他在晶体管理论化方面的贡献很重要。但正是这八名叛逆的年轻工程师抛弃了他的公司,以及德州仪器 (Texas Instruments) 的一个类似团队,他们将 Shockley 的晶体管变成了一种有用的产品——芯片——并将它们卖给了美国军方,同时学习了如何大规模生产它们。凭借这些能力,Fairchild 和 TI 进入了 1960 年代中期,面临着新的挑战:将芯片转变为大众市场产品。

Shockley, who was widely recognized as one of the greatest theoretical physicists of his generation, eventually abandoned his effort to make a fortune and get his name in the Wall Street Journal. His contribution in theorizing the transistor was important. But it was the traitorous eight young engineers who abandoned his company, as well as a similar group at Texas Instruments, who turned Shockley’s transistors into a useful product—chips—and sold them to the U.S. military while learning how to mass-produce them. Armed with these capabilities, Fairchild and TI entered the mid-1960s with a new challenge: turning chips into a mass market product.

第 6 章 “我……想要……变得……富有”

CHAPTER 6 “I… WANT… TO… GET… RICH”

引导阿波罗宇宙飞船和民兵 II 导弹的计算机为美国的集成电路产业提供了最初的升空。到 1960 年代中期,美国军方在从卫星到声纳的各种武器中部署芯片,鱼雷到遥测系统。Bob Noyce 知道军事和太空计划对于 Fairchild 早期的成功至关重要,他在 1965 年承认军事和太空应用将使用“今年生产的超过 95% 的电路”。但他总是为他的芯片设想一个更大的民用市场,尽管在 1960 年代初期还没有这样的市场。他必须创造它,这意味着要与军方保持一定距离,以便他——而不是五角大楼——确定仙童的研发重点。诺伊斯拒绝了大多数军事研究合同,估计飞兆半导体的研发预算从不依赖国防部超过 4%。诺伊斯自信地解释说:“世界上很少有研究主任能真正胜任”评估仙童工作的工作,“而且他们是陆军中的职业军官并不常见。”

The computers that guided the Apollo spacecraft and the Minuteman II missile provided the initial liftoff for America’s integrated circuit industry. By the mid-1960s, the U.S. military was deploying chips in weaponry of all types, from satellites to sonar, torpedoes to telemetry systems. Bob Noyce knew that military and space programs were crucial for Fairchild’s early success, admitting in 1965 that military and space applications would use “over 95% of the circuits produced this year.” But he always envisioned an even larger civilian market for his chips, though in the early 1960s no such market existed. He would have to create it, which meant keeping the military at arm’s length so that he—not the Pentagon—set Fairchild’s R&D priorities. Noyce declined most military research contracts, estimating that Fairchild never relied on the Defense Department for more than 4 percent of its R&D budget. “There are very few research directors anywhere in the world who are really adequate to the job” of assessing Fairchild’s work, Noyce explained confidently, “and they are not often career officers in the Army.”

诺伊斯刚从研究生院毕业时就曾在 Philco 工作,这是一家拥有大型国防部门的东海岸无线电制造商,他曾经历过政府指导的研发工作。“研究方向是由能力较弱的人决定的,”诺伊斯回忆说,抱怨他浪费时间为军队写进度报告。现在他正在经营一家由信托基金继承人创办的公司 Fairchild,他可以灵活地将军队视为客户而不是老板。他选择将 Fairchild 的大部分研发目标不是军用,而是大众市场产品。他推断,火箭或卫星中使用的大多数芯片也必须有民用用途。第一个为商业市场生产的集成电路,用于Zenith 助听器,最初是为 NASA 卫星设计的。挑战在于制造平民能买得起的芯片。军方付出了高昂的代价,但消费者对价格很敏感。然而,仍然令人着迷的是,民用市场甚至比冷战五角大楼臃肿的预算还要大得多。“向政府出售研发就像把你的风险资本存入储蓄账户,”诺伊斯宣称。“冒险就是冒险;你愿意冒险。”

Noyce had experienced government-directed R&D while fresh out of graduate school when he worked for Philco, an East Coast radio manufacturer with a big defense unit. “The direction of the research was being determined by people less competent,” Noyce recalled, complaining about the time he wasted writing progress reports for the military. Now that he was running Fairchild, a company seeded by a trust-fund heir, he had flexibility to treat the military as a customer rather than a boss. He chose to target much of Fairchild’s R&D not at the military, but at mass market products. Most of the chips used in rockets or satellites must have civilian uses, too, he reasoned. The first integrated circuit produced for commercial markets, used in a Zenith hearing aid, had initially been designed for a NASA satellite. The challenge would be making chips that civilians could afford. The military paid top dollar, but consumers were price sensitive. What remained tantalizing, though, was that the civilian market was far larger than even the bloated budgets of the Cold War Pentagon. “Selling R&D to the government was like taking your venture capital and putting it into a savings account,” Noyce declared. “Venturing is venturing; you want to take the risk.”

在帕洛阿尔托,仙童半导体被五角大楼的供应商所包围,从航空航天到弹药,从无线电到雷达。尽管军方从 Fairchild 购买了芯片,但国防部更愿意与大型官僚机构合作,而不是与灵活的初创公司合作。因此,五角大楼低估了仙童和其他半导体初创公司改造电子产品的速度。国防部在 1950 年代后期的一项评估中称赞无线电巨头 RCA 正在进行“最雄心勃勃的微型化计划”,同时不屑一顾地指出,仙童公司只有两名科学家从事公司领先的电路计划。国防承包商洛克希德马丁公司在帕洛阿尔托附近有一个研究设施,在他们的微系统电子部门有超过 50 名科学家,洛克希德遥遥领先。

In Palo Alto, Fairchild Semiconductor was surrounded by firms that supplied the Pentagon, from aerospace to ammunition, radio to radar. Though the military bought chips from Fairchild, the Defense Department was more comfortable working with big bureaucracies than nimble startups. As a result, the Pentagon underestimated the speed at which Fairchild and other semiconductor startups would transform electronics. A Defense Department assessment from the late 1950s had praised radio giant RCA for having “the most ambitious microminiaturization program underway” while dismissively noting that Fairchild had only two scientists working on the company’s leading circuit program. Defense contractor Lockheed Martin, which had a research facility just down the road in Palo Alto, had over fifty scientists in their microsystem electronics division, the Defense Department reported, implying that Lockheed was far ahead.

然而,在 Gordon Moore 的指导下,Fairchild 的研发团队不仅设计了新技术,还开辟了新的民用市场。1965 年,摩尔被《电子》杂志要求写一篇关于集成电路未来的短文。他预测,至少在接下来的十年中,Fairchild 每年都会使硅芯片上可以容纳的组件数量增加一倍。如果是这样,到 1975 年,集成电路将有 65,000 个微型晶体管刻在其中,这不仅可以提高计算能力,而且每个晶体管的价格也会降低。随着成本下降,用户数量将会增加。这种对计算能力指数级增长的预测很快后来被称为摩尔定律。这是本世纪最伟大的技术预测。

However, it was Fairchild’s R&D team that, under Gordon Moore’s direction, not only devised new technology but opened new civilian markets as well. In 1965, Moore was asked by Electronics magazine to write a short article on the future of integrated circuits. He predicted that every year for at least the next decade, Fairchild would double the number of components that could fit on a silicon chip. If so, by 1975, integrated circuits would have sixty-five thousand tiny transistors carved into them, creating not only more computing power but also lower prices per transistor. As costs fell, the number of users would grow. This forecast of exponential growth in computing power soon came to be known as Moore’s Law. It was the greatest technological prediction of the century.

摩尔意识到,如果每个芯片上的计算能力继续呈指数级增长,集成电路将彻底改变社会,远远超出火箭和雷达。1965 年,国防资金仍然购买了当年生产的所有集成电路的 72%。但是,军方要求的功能在商业应用中也很有用。“小型化和坚固性”,一份电子出版物宣称,“意味着好生意。” 国防承包商主要认为芯片是一种可以替代所有军事系统中旧电子设备的产品。在仙童,诺伊斯和摩尔已经梦想着个人电脑和手机。

If the computing power on each chip continued to grow exponentially, Moore realized, the integrated circuit would revolutionize society far beyond rockets and radars. In 1965, defense dollars still bought 72 percent of all integrated circuits produced that year. However, the features the military demanded were useful in business applications, too. “Miniaturization and ruggedness,” one electronics publication declared, “means good business.” Defense contractors thought about chips mostly as a product that could replace older electronics in all the military’s systems. At Fairchild, Noyce and Moore were already dreaming of personal computers and mobile phones.

当美国国防部长罗伯特麦克纳马拉在 1960 年代初期改革军事采购以削减成本时,导致了电子行业中一些人所说的“麦克纳马拉萧条”,飞兆半导体对民用芯片的愿景似乎具有先见之明。该公司是第一家为民用客户提供完整的现成集成电路产品线的公司。诺伊斯也大幅降价,赌注这将大大扩大民用芯片市场。在 1960 年代中期,以前售价 20 美元的仙童芯片被削减到 2 美元。有时Fairchild 甚至以低于制造成本的价格出售产品,希望能说服更多客户尝试。

When U.S. defense secretary Robert McNamara reformed military procurement to cut costs in the early 1960s, causing what some in the electronics industry called the “McNamara Depression,” Fairchild’s vision of chips for civilians seemed prescient. The company was the first to offer a full product line of off-the-shelf integrated circuits for civilian customers. Noyce slashed prices, too, gambling that this would drastically expand the civilian market for chips. In the mid-1960s, Fairchild chips that previously sold for $20 were cut to $2. At times Fairchild even sold products below manufacturing cost, hoping to convince more customers to try them.

由于价格下跌,Fairchild 开始赢得私营部门的主要合同。美国计算机年销量从 1957 年的 1,000 台增长到十年后的 18,700 台。到 1960 年代中期,几乎所有这些计算机都依赖于集成电路。1966 年,计算机公司 Burroughs 从 Fairchild 订购了 2000 万个芯片——是当时的 20 多倍。阿波罗计划消耗。到 1968 年,计算机行业购买的芯片数量与军用芯片一样多。Fairchild 芯片服务于这个计算机市场的 80%。鲍勃诺伊斯的降价获得了回报,为民用计算机开辟了一个新市场,这将推动未来几十年的芯片销售。摩尔后来辩称,诺伊斯的降价与仙童集成电路中的技术一样大的创新。

Thanks to falling prices, Fairchild began winning major contracts in the private sector. Annual U.S. computer sales grew from 1,000 in 1957 to 18,700 a decade later. By the mid-1960s, almost all these computers relied on integrated circuits. In 1966, Burroughs, a computer firm, ordered 20 million chips from Fairchild—more than twenty times what the Apollo program consumed. By 1968, the computer industry was buying as many chips as the military. Fairchild chips served 80 percent of this computer market. Bob Noyce’s price cuts had paid off, opening a new market for civilian computers that would drive chip sales for decades to come. Moore later argued that Noyce’s price cuts were as big an innovation as the technology inside Fairchild’s integrated circuits.

到 1960 年代末,经过十年的发展,阿波罗 11 号终于准备好使用其仙童驱动的导航计算机将第一个人类送上月球。加州圣克拉拉谷的半导体工程师从太空竞赛中受益匪浅,这提供了一个重要的早期客户。然而,到第一次登月时,硅谷的工程师对国防和太空合同的依赖程度已经大大降低。现在他们专注于更世俗的关注。芯片市场蓬勃发展。Fairchild 的成功已经激发了几位顶级员工投奔竞争对手的芯片制造商。风险投资资金正涌入那些不专注于火箭而是专注于企业计算机的初创公司。

By the end of the 1960s, after a decade of development, Apollo 11 was finally ready to use its Fairchild-powered guidance computer to carry the first human to the moon. The semiconductor engineers in California’s Santa Clara Valley had benefitted immensely from the space race, which provided a crucial early customer. Yet by the time of the first lunar landing, Silicon Valley’s engineers had become far less dependent on defense and space contracts. Now they were focused on more earthly concerns. The chip market was booming. Fairchild’s success had already inspired several top employees to defect to competing chipmakers. Venture capital funding was pouring into startups that focused not on rockets but on corporate computers.

然而,Fairchild 仍然为一位东海岸的千万富翁所拥有,他向员工支付了丰厚的报酬,但拒绝给予他们股票期权,将放弃股权的想法视为一种形式“爬行的社会主义”。最终,即使是仙童的联合创始人之一诺伊斯也开始怀疑他在公司是否有未来。很快,每个人都开始寻找出口。原因很明显。除了新的科学发现和新的制造工艺之外,这种造成财务损失的能力是推动摩尔定律向前发展的根本力量。正如 Fairchild 的一名员工在离开公司时填写的离职问卷中所说:“我……想要……变得……变得……富有。”

Fairchild, however, was still owned by an East Coast multimillionaire who paid his employees well but refused to give them stock options, viewing the idea of giving away equity as a form of “creeping socialism.” Eventually, even Noyce, one of Fairchild’s cofounders, began wondering whether he had a future at the firm. Soon everyone began looking for the exit. The reason was obvious. Alongside new scientific discoveries and new manufacturing processes, this ability to make a financial killing was the fundamental force driving forward Moore’s Law. As one of Fairchild’s employees put it in the exit questionnaire he filled out when leaving the company: “I… WANT… TO… GET… RICH.”

第二部分 美国世界的电路

PART II THE CIRCUITRY OF THE AMERICAN WORLD

第七章 苏联硅谷

CHAPTER 7 Soviet Silicon Valley

Bob Noyce 在 Fairchild Semiconductor 发明集成电路几个月后一位不速之客来到了帕洛阿尔托。1959 年秋天,在人造卫星首次绕地球运行两年后,来自苏联的半导体工程师阿纳托利·特鲁特科搬进了斯坦福大学的一个名为克罗瑟斯纪念馆的宿舍。尽管冷战竞争已接近高潮,但两个超级大国已同意开始学生交流,而特鲁特科是少数被苏联选中并经美国国务院审查的学生之一。他在斯坦福大学与美国顶尖科学家一起研究美国最先进的技术。他甚至参加了威廉·肖克利(William Shockley)的讲座,后者已经放弃了自己的创业公司,现在是大学的教授。一节课后,特鲁特科请诺贝尔奖得主在他的巨著《半导体中的电子和空穴》的副本上签名. “致阿纳托尔,”肖克利签了字,然后对这位年轻的科学家大吼大叫,抱怨苏联拒绝为教科书的俄文翻译支付版税。

A couple months after Bob Noyce invented his integrated circuit at Fairchild Semiconductor, an unexpected visitor arrived in Palo Alto. In fall 1959, two years after Sputnik first orbited the earth, Anatoly Trutko, a semiconductor engineer from the Soviet Union, moved into a Stanford University dormitory called Crothers Memorial Hall. Though Cold War competition was near its peak, the two superpowers had agreed to begin student exchanges, and Trutko was one of a handful of students selected by the USSR and vetted by the U.S. State Department. He spent his year at Stanford studying America’s most advanced technology with the country’s leading scientists. He even attended lectures given by William Shockley, who’d abandoned his startup and was now a professor at the university. After one class, Trutko asked the Nobel Prize winner to sign a copy of his magnum opus Electrons and Holes in Semiconductors. “To Anatole,” Shockley signed, before barking at the young scientist with complaints that the Soviet Union refused to pay royalties for the textbook’s Russian translation.

鉴于美国担心苏联在科学和技术方面正在迎头赶上,美国决定让特鲁特科这样的苏联科学家在斯坦福大学研究半导体是令人惊讶的。然而,每个国家的电子行业都越来越倾向于硅谷,它完全设定了创新的标准和步伐,以至于其他国家世界别无选择,只能跟随——即使是美国的对手。苏联没有向肖克利支付版税,但他们了解半导体的价值,在肖克利的教科书出版两年后就将其翻译成俄文。早在 1956 年,美国的间谍就被命令收购苏联的半导体设备,以测试它们的质量并跟踪它们的改进。中情局 1959 年的一份报告发现,美国在生产晶体管的质量和数量上仅领先苏联两到四年。至少有几名早期的苏联交换生是克格勃特工——当时被怀疑,但直到几十年后才被证实——在学生交流和苏联国防工业目标之间建立了密切的联系。

America’s decision to let Soviet scientists like Trutko study semiconductors at Stanford was surprising, given U.S. fears that the Soviet Union was catching up in science and technology. Yet every country’s electronics industry was increasingly oriented toward Silicon Valley, which so totally set the standard and pace of innovation that the rest of the world had no choice but to follow—even America’s adversaries. The Soviets didn’t pay Shockley royalties, but they understood the value of semiconductors, translating Shockley’s textbook into Russian just two years after it was published. As early as 1956, America’s spies had been ordered to acquire Soviet semiconductor devices to test their quality and track their improvements. A CIA report in 1959 found that America was only two to four years ahead of the Soviets in quality and quantity of transistors produced. At least several of the early Soviet exchange students were KGB agents—suspected at the time, but not confirmed until decades later—forging an intimate connection between student exchanges and Soviet defense industrial goals.

就像五角大楼一样,克里姆林宫意识到晶体管和集成电路将改变制造、计算和军事力量。从 1950 年代后期开始,苏联在全国建立了新的半导体设施,并指派了最聪明的科学家来建立这个新产业。对于像 Yuri Osokin 这样雄心勃勃的年轻工程师来说,很难想象有比这更令人兴奋的任务了。奥索金童年的大部分时间都在中国度过,他的父亲在黄海沿岸大连市的一家苏联军队医院工作。从年轻时起,奥索金就因其对地理和名人生日等方面的百科全书式记忆而脱颖而出。完成学业后,他进入莫斯科一所顶尖学术机构,专攻半导体。

Just like the Pentagon, the Kremlin realized that transistors and integrated circuits would transform manufacturing, computing, and military power. Beginning in the late 1950s, the USSR established new semiconductor facilities across the country and assigned its smartest scientists to build this new industry. For an ambitious young engineer like Yuri Osokin it was hard to imagine a more exciting assignment. Osokin had spent much of his childhood in China, where his father worked in a Soviet military hospital in the city of Dalian, on the shores of the Yellow Sea. From his youth, Osokin stood out for his encyclopedic memory for things like geography and the birthdays of famous people. After finishing school, he won entrance to a top academic institute in Moscow and specialized in semiconductors.

奥索金很快被分配到里加的一家半导体工厂,配备了来自该国最好大学的应届毕业生,并受命为苏联太空计划和军队制造半导体设备。工厂主任委托奥索金在同一块锗上建立一个包含多个组件的电路,这是苏联以前没有人做过的事情。他于 1962 年制作了他的原型集成电路。奥索金和他的同事们知道他们处于苏联科学的最前沿。他们白天在实验室里修修补补,晚上则讨论固态物理学理论,奥索金偶尔会拿出他的吉他来陪伴他的同事在唱歌。他们很年轻,他们的工作令人兴奋,苏联的科学正在崛起,苏联的几颗人造卫星在头顶运行,肉眼可见每当奥索金放下吉他,抬头仰望夜空。

Osokin was soon assigned to a semiconductor plant in Riga, staffed with fresh graduates from the country’s best universities, and ordered to build semiconductor devices for the Soviet space program and the military. Osokin was tasked by the factory’s director to build a circuit with multiple components on the same piece of germanium, something no one in the Soviet Union had previously done. He produced his prototype integrated circuit in 1962. Osokin and his colleagues knew they were at the cutting edge of Soviet science. They spent their days tinkering in labs and their evenings debating the theory of solid-state physics, with Osokin occasionally breaking out his guitar to accompany his colleagues in song. They were young, their work was exciting, Soviet science was rising, and several of the USSR’s Sputnik satellites were orbiting overhead, visible to the naked eye whenever Osokin put down his guitar and looked up into the night sky.

苏联领导人尼基塔赫鲁晓夫致力于在各个领域超越美国,从玉米生产到卫星发射。赫鲁晓夫本人在集体农场比在电子实验室更自在。他对技术一窍不通,却痴迷于“赶超”美国的概念,正如他一再承诺的那样。苏联国家无线电电子委员会第一副主席亚历山大·肖金意识到,赫鲁晓夫与美国竞争的冲动可以用来赢得更多的微电子投资。“想象一下,尼基塔·谢尔盖耶维奇,”肖金有一天对苏联领导人说,“可以制造一台电视机烟盒大小。” 这就是苏联硅的承诺。“赶超”美国似乎是一个真正的可能性。与苏联追赶美国的另一个领域——核武器一样,苏联有一个秘密武器:一个间谍圈。

Soviet leader Nikita Khrushchev was committed to outcompeting the United States in every sphere, from corn production to satellite launches. Khrushchev himself was more comfortable on collective farms than in electronics labs. He understood nothing about technology but was obsessed with the notion of “catching up and overtaking” the United States, as he repeatedly promised to do. Alexander Shokin, first deputy chairman of the Soviet State Committee on Radioelectronics, realized Khrushchev’s urge to compete with the United States could be used to win more investment in microelectronics. “Imagine, Nikita Sergeyevich,” Shokin told the Soviet leader one day, “that a TV can be made the size of a cigarette box.” Such was the promise of Soviet silicon. “Catching up and overtaking” the United States seemed like a real possibility. As with another sphere where the Soviets had caught up to the United States—nuclear weapons—the USSR had a secret weapon: a spy ring.

乔尔·巴尔是两名俄罗斯犹太人的儿子,他们移民到美国以逃避沙皇的压迫。巴尔在布鲁克林贫困长大,后来进入纽约城市学院学习电气工程。作为一名学生,他结识了一群共产党人,并发现自己同情他们对资本主义的批评以及他们认为苏联最有能力对抗纳粹的论点。通过共产党人脉,他认识了阿尔弗雷德·萨兰特(Alfred Sarant),他是一位电气工程师,也是共青团的成员。他们将度过余生,共同推动共产主义事业。

Joel Barr was the son of two Russian Jews who immigrated to the U.S. to flee tsarist oppression. Barr grew up in poverty in Brooklyn before winning admission to the City College of New York to study electrical engineering. As a student, he fell in with a group of Communists and found himself sympathizing with their critique of capitalism and their argument that the Soviet Union was best placed to stand up to the Nazis. Via Communist Party contacts, he was introduced to Alfred Sarant, a fellow electrical engineer and member of the Young Communist League. They’d spend the remainder of their lives working together to further the Communist cause.

在 1930 年代,巴尔和萨兰特被整合到臭名昭著的冷战间谍朱利叶斯·罗森伯格领导的间谍团伙中。在 1940 年代,Barr 和 Sarant 在美国两家领先的技术公司 Western Electric 和 Sperry Gyroscope 从事机密雷达和其他军事系统的工作。与 Rosenberg 环中的其他人不同,Barr 和萨兰特不掌握核武器的秘密,但他们对新武器系统中的电子设备有了深入的了解。在 1940 年代后期,随着联邦调查局开始解开克格勃在美国的间谍网络,罗森伯格与他的妻子埃塞尔一起被审判并被判处电刑死刑。在联邦调查局抓住他们之前,萨兰特和巴尔逃离了这个国家,最终到达了苏联。

In the 1930s, Barr and Sarant were integrated into an espionage ring led by Julius Rosenberg, the infamous Cold War spy. During the 1940s, Barr and Sarant worked on classified radars and other military systems at Western Electric and Sperry Gyroscope, two leading American technology firms. Unlike others in the Rosenberg ring, Barr and Sarant didn’t possess nuclear weapons secrets, but they had gained intimate knowledge about the electronics in new weapons systems. In the late 1940s, as the FBI began unraveling the KGB’s spy networks in the U.S., Rosenberg was tried and sentenced to death by electrocution alongside his wife, Ethel. Before the FBI could catch them, Sarant and Barr fled the country, eventually reaching the Soviet Union.

当他们到达时,他们告诉克格勃处理人员他们想制造世界上最先进的计算机。巴尔和萨兰特不是计算机专家,苏联的其他人也不是。他们作为间谍的身份本身就是一种备受推崇的凭证,他们的光环使他们能够获得资源。在 1950 年代后期,Barr 和 Sarant 开始建造他们的第一台计算机,称为 UM——俄语中“头脑”的意思。他们的工作引起了管理苏联电子工业的官僚 Shokin 的注意,他们与他合作说服赫鲁晓夫,苏联需要一个致力于生产半导体的整个城市,并拥有自己的研究人员、工程师、实验室和生产设施。

When they arrived, they told KGB handlers they wanted to build the world’s most advanced computers. Barr and Sarant weren’t experts in computers, but nor was anyone else in the Soviet Union. Their status as spies was, in itself, a much admired credential, and their aura gave them access to resources. In the late 1950s, Barr and Sarant began building their first computer, called UM—the Russian word for “mind.” Their work attracted the attention of Shokin, the bureaucrat who managed the Soviet electronics industry, and they partnered with him to convince Khrushchev that the USSR needed an entire city devoted to producing semiconductors, with its own researchers, engineers, labs, and production facilities. Even before the towns on the peninsula south of San Francisco had become known as Silicon Valley—a term that wasn’t coined until 1971—Barr and Sarant had dreamt up their own version in a Moscow suburb.

为了说服赫鲁晓夫资助这座新的科学城,肖金安排苏联领导人访问列宁格勒的第二电子工业特别设计局。在庞大而官僚的名字背后——苏联人从不擅长营销——是一家处于苏联电子技术前沿的研究所。设计局花了数周时间为赫鲁晓夫的来访做准备,前一天举行了彩排,以确保一切按计划进行。1962年5月4日,赫鲁晓夫抵达。为了迎接这位苏联领导人,萨兰特身着深色西装,与他浓密的眉毛和精心修剪的小胡子颜色相配。巴尔紧张地站在萨兰特身边,他光秃的脑袋上架着一副金属眼镜。在萨兰特的带领下,两位前间谍向赫鲁晓夫展示了苏联微电子的成就。赫鲁晓夫测试了一个适合他耳朵和玩弄的微型收音机用一台可以打印出他名字的简单电脑。半导体设备很快将用于航天器、工业、政府、飞机——甚至“用于制造核导弹防护罩”,萨兰特自信地告诉赫鲁晓夫。然后他和巴尔把赫鲁晓夫带到一个画架前,画架上画着一座专门生产半导体设备的未来城市,中心有一座巨大的 52 层摩天大楼。

To convince Khrushchev to fund this new city of science, Shokin arranged for the Soviet leader to visit Special Design Bureau of the Electronics Industry #2 in Leningrad. Behind the bulky, bureaucratic name—the Soviets never excelled at marketing—was an institute at the cutting edge of Soviet electronics. The Design Bureau spent weeks preparing for Khrushchev’s visit, holding a dress rehearsal the day before to ensure that everything went according to plan. On May 4, 1962, Khrushchev arrived. To welcome the Soviet leader, Sarant dressed in a dark suit matching the color of his bushy eyebrows and carefully trimmed mustache. Barr stood nervously to Sarant’s side, wiry glasses perched on his balding head. With Sarant in the lead, the two former spies showed Khrushchev the accomplishments of Soviet microelectronics. Khrushchev tested a tiny radio that fit in his ear and toyed with a simple computer that could print out his name. Semiconductor devices would soon be used in spacecraft, industry, government, aircraft—even “for the creation of a nuclear missile shield,” Sarant confidently told Khrushchev. Then he and Barr led Khrushchev to an easel with pictures of a futuristic city devoted exclusively to producing semiconductor devices, with a vast fifty-two-story skyscraper at its center.

赫鲁晓夫痴迷于宏大的项目,尤其是那些他可以宣称功劳的项目,因此他热情地支持为半导体建造苏联城市的想法。他以熊抱的方式拥抱了巴尔和萨兰特,承诺全力支持。几个月后,苏联政府批准了在莫斯科郊区建设半导体城的计划。“微电子学是一个机械大脑,”赫鲁晓夫向他的苏联领导人解释道。“这是我们的未来。”

Khrushchev was enamored of grand projects, especially those that he could claim credit for, so he enthusiastically endorsed the idea of building a Soviet city for semiconductors. He embraced Barr and Sarant in a bear hug, promising his full support. Several months later, the Soviet government approved plans to build a semiconductor city in the outskirts of Moscow. “Microelectronics is a mechanical brain,” Khrushchev explained to his fellow Soviet leaders. “It is our future.”

苏联很快在泽列诺格勒市破土动工,该市在俄语中意为“绿色城市”——事实上,它被设计成一个科学天堂。Shokin 希望它成为一个完美的科学解决方案,拥有研究实验室和生产设施,加上学校、日托中心、电影院、图书馆和医院——半导体工程师可能需要的一切。靠近中心的是一所大学,即莫斯科电子技术学院,其砖外墙模仿了英国和美国的大学校园。从外面看,它就像硅谷,只是少了一点阳光。

The USSR soon broke ground on the city of Zelenograd, the Russian word for “green city”—and, indeed, it was designed to be a scientific paradise. Shokin wanted it to be a perfect scientific settlement, with research laboratories and production facilities, plus schools, day cares, movie theaters, libraries, and a hospital—everything a semiconductor engineer could need. Near the center was a university, the Moscow Institute of Electronic Technology, with a brick façade modeled on English and American college campuses. From the outside, it seemed just like Silicon Valley, only a little less sunny.

第 8 章 “复制它”

CHAPTER 8 “Copy It”

一个就在尼基塔·赫鲁晓夫宣布支持建造泽列诺格勒的同时,一位名叫鲍里斯·马林的苏联学生从宾夕法尼亚州学习了一年回来,他的行李箱里放着一个小设备——德州仪器公司的 SN-51,这是美国最早出售的集成电路之一。美国。马林是一个瘦削的男人,有着深色的头发和深陷的眼睛,他是苏联半导体设备方面的主要专家之一。他认为自己是一名科学家,而不是间谍。然而,负责苏联微电子的官僚亚历山大·肖金认为,SN-51 是苏联必须通过任何必要手段获得的设备。肖金把马林和其他工程师叫到他的办公室,把芯片放在显微镜下,透过镜头观察。“复制它,”他命令他们,“一对一,没有任何偏差。我给你三个月的时间。”

Around the same time that Nikita Khrushchev declared his support for building Zelenograd, a Soviet student named Boris Malin returned from a year studying in Pennsylvania with a small device in his luggage—a Texas Instruments SN-51, one of the first integrated circuits sold in the United States. A thin man with dark hair and deep-set eyes, Malin was one of the Soviet Union’s leading experts on semiconductor devices. He saw himself as a scientist, not a spy. Yet Alexander Shokin, the bureaucrat in charge of Soviet microelectronics, believed the SN-51 was a device the Soviet Union must acquire by any means necessary. Shokin called Malin and a group of other engineers into his office, placed the chip under his microscope, and peered through the lens. “Copy it,” he ordered them, “one-for-one, without any deviations. I’ll give you three months.”

苏联科学家对他们只是抄袭外国进步的说法感到愤怒。他们的科学理解与美国的化学家和物理学家一样先进。在美国的苏联交换生报告说,他们从威廉·肖克利的讲座中学到的东西很少,他们无法在莫斯科学习。事实上,苏联拥有一些世界领先的理论物理学家。当杰克·基尔比(Jack Kilby)最终因发明集成电路而在 2000 年获得诺贝尔物理学奖(当时是集成电路的共同发明者,Bob Noyce,已经去世),他与一位名叫 Zhores Alferov 的俄罗斯科学家分享了这个奖项,他在 1960 年代就半导体设备产生光的方式进行了基础研究。1957 年人造卫星的发射、1961 年尤里·加加林的首次太空飞行以及 1962 年奥索金集成电路的制造为苏联正在成为科学超级大国提供了无可争辩的证据。甚至中央情报局也认为苏联的微电子工业正在迅速追赶。

Soviet scientists reacted angrily to the suggestion they were simply copying foreign advances. Their scientific understanding was as advanced as that of America’s chemists and physicists. Soviet exchange students in the U.S. reported learning little from lectures by William Shockley that they couldn’t have studied in Moscow. Indeed, the USSR had some of the world’s leading theoretical physicists. When Jack Kilby was finally awarded the Nobel Prize in Physics in 2000 for inventing the integrated circuit (by then the co-inventor of the integrated circuit, Bob Noyce, had died), he shared the prize with a Russian scientist named Zhores Alferov, who’d conducted fundamental research in the 1960s on ways semiconductor devices could produce light. The launch of Sputnik in 1957, the first space flight of Yuri Gagarin in 1961, and the fabrication of Osokin’s integrated circuit in 1962 provided incontrovertible evidence that the Soviet Union was becoming a scientific superpower. Even the CIA thought the Soviet microelectronics industry was catching up rapidly.

然而,Shokin 的“复制”策略存在根本缺陷。复制在制造核武器方面很有效,因为美国和苏联在整个冷战期间只制造了数万枚核武器。然而,在美国,TI 和 Fairchild 已经在学习如何大规模生产芯片。扩大生产规模的关键是可靠性,这是美国芯片制造商如 Morris Chang 和 Andy Grove 在 1960 年代所关注的挑战。与苏联同行不同,他们可以利用其他公司制造先进光学、化学品、纯化材料和其他生产机械的专业知识。如果没有美国公司可以提供帮助,Fairchild 和 TI 可以求助于德国、法国或英国,它们都拥有自己的先进产业。

Shokin’s “copy it” strategy was fundamentally flawed, however. Copying worked in building nuclear weapons, because the U.S. and the USSR built only tens of thousands of nukes over the entire Cold War. In the U.S., however, TI and Fairchild were already learning how to mass-produce chips. The key to scaling production was reliability, a challenge that American chipmakers like Morris Chang and Andy Grove fixated on during the 1960s. Unlike their Soviet counterparts, they could draw on the expertise of other companies making advanced optics, chemicals, purified materials, and other production machinery. If no American companies could help, Fairchild and TI could turn to Germany, France, or Britain, each of which had advanced industries of their own.

苏联大量生产煤炭和钢铁,但几乎在所有类型的先进制造业中都落后。苏联在数量上表现出色,但在质量或纯度上并不出色,这两者对于大批量芯片制造至关重要。此外,西方盟国禁止通过一个名为 COCOM 的组织向共产主义国家转让包括半导体组件在内的许多先进技术。苏联人通常可以利用中立的奥地利或瑞士的空壳公司绕过 COCOM 的限制,但这种途径很难大规模使用。因此,苏联的半导体设施经常不得不使用不太复杂的机器和不太纯净的材料,因此生产的工作芯片要少得多。

The Soviet Union churned out coal and steel in vast quantities but lagged in nearly every type of advanced manufacturing. The USSR excelled in quantity but not in quality or purity, both of which were crucial to high-volume chipmaking. Moreover, the Western allies prohibited the transfer of many advanced technologies, including semiconductor components, to Communist countries via an organization called COCOM. The Soviets could often bypass COCOM restrictions using shell companies in neutral Austria or Switzerland, but this pathway was hard to use on a large-scale basis. So Soviet semiconductor facilities regularly had to work with machinery that was less sophisticated and with materials that were less pure, producing far fewer working chips as a result.

到目前为止,间谍活动只能得到 Shokin 和他的工程师。仅仅偷一块芯片并不能解释它是如何制作的,就像偷蛋糕不能解释它是如何烘烤的一样。薯片的配方已经异常复杂。在斯坦福大学跟随肖克利学习的交换生可以成为聪明的物理学家,但像安迪·格罗夫或玛丽·安妮·波特这样的工程师知道某些化学物质需要在什么温度下加热,或者光刻胶应该暴露在光线下多长时间。制造芯片过程的每一步都涉及专业知识,而这些专业知识很少在特定公司之外共享。这种类型的专有技术通常甚至没有被记录下来。苏联间谍是业内最优秀的间谍之一,但半导体生产过程需要更多的细节和知识,即使是最有能力的间谍也无法窃取。

Spying could only get Shokin and his engineers so far. Simply stealing a chip didn’t explain how it was made, just as stealing a cake can’t explain how it was baked. The recipe for chips was already extraordinarily complicated. Foreign exchange students studying with Shockley at Stanford could become smart physicists, but it was engineers like Andy Grove or Mary Anne Potter who knew at what temperature certain chemicals needed to be heated, or how long photoresists should be exposed to light. Every step of the process of making chips involved specialized knowledge that was rarely shared outside of a specific company. This type of know-how was often not even written down. Soviet spies were among the best in the business, but the semiconductor production process required more details and knowledge than even the most capable agent could steal.

此外,按照摩尔定律中规定的速率,尖端在不断变化。即使苏联人设法复制设计、获取材料和机械并复制生产过程,这也需要时间。TI 和 Fairchild 每年都在推出具有更多晶体管的新设计。到 1960 年代中期,最早的集成电路已成为旧闻,体积太大且耗电量大,价值不高。与几乎任何其他任何类型的技术相比,半导体技术都在飞速发展。晶体管的尺寸和它们的能耗正在缩小,而可以装在一平方英寸硅上的计算能力大约每两年翻一番。没有其他技术发展得如此之快——因此,没有其他领域的窃取去年的设计是一种如此绝望的策略。

Moreover, the cutting edge was constantly changing, per the rate set out in Moore’s Law. Even if the Soviets managed to copy a design, acquire the materials and machinery, and replicate the production process, this took time. TI and Fairchild were introducing new designs with more transistors every year. By the mid-1960s, the earliest integrated circuits were old news, too big and power-hungry to be very valuable. Compared to almost any other any type of technology, semiconductor technology was racing forward. The size of transistors and their energy consumption was shrinking, while the computing power that could be packed on a square inch of silicon roughly doubled every two years. No other technology moved so quickly—so there was no other sector in which stealing last year’s design was such a hopeless strategy.

苏联领导人从未理解“复制”战略如何使他们陷入落后。整个苏联半导体部门的运作就像一个国防承包商——秘密的、自上而下的、面向军事系统的、执行订单而几乎没有创造力的空间。肖金部长“严密控制”了复制过程,他的一名下属回忆道。复制在苏联的半导体行业中确实是根深蒂固的,一些芯片制造机器使用英寸而不是厘米来更好地复制美国的设计,尽管苏联其他地区使用公制。得益于“复制”战略,苏联在晶体管技术方面落后美国数年,但从未赶上。

Soviet leaders never comprehended how the “copy it” strategy condemned them to backwardness. The entire Soviet semiconductor sector functioned like a defense contractor—secretive, top-down, oriented toward military systems, fulfilling orders with little scope for creativity. The copying process was “tightly controlled” by Minister Shokin, one of his subordinates remembered. Copying was literally hardwired into the Soviet semiconductor industry, with some chipmaking machinery using inches rather than centimeters to better replicate American designs, even though the rest of the USSR used the metric system. Thanks to the “copy it” strategy, the USSR started several years behind the U.S. in transistor technology and never caught up.

泽列诺格勒可能看起来就像没有阳光的硅谷。它拥有该国最好的科学家和被盗的秘密。然而,两国的半导体系统截然不同。硅谷的初创公司创始人跳槽并获得了“在工厂车间”的实际经验,而肖金则在莫斯科的部长办公桌上发号施令。与此同时,尤里·奥索金在里加默默无闻地生活着,受到同事的高度尊重,但无法与任何缺乏安全许可的人谈论他的发明。年轻的苏联学生没有攻读电气工程学位,他们想要像奥索金一样,因为没有人知道他的存在。职业发展需要成为一个更好的官僚,而不是设计新产品或寻找新市场。在高度关注军事生产的情况下,民用产品始终是事后的想法。

Zelenograd might have seemed like Silicon Valley without the sunshine. It had the country’s best scientists and stolen secrets. Yet the two countries’ semiconductor systems couldn’t have been more different. Whereas Silicon Valley’s startup founders job-hopped and gained practical “on the factory floor” experience, Shokin called the shots from his ministerial desk in Moscow. Yuri Osokin, meanwhile, lived in obscurity in Riga, highly respected by his colleagues but unable to speak about his invention with anyone who lacked a security clearance. Young Soviet students didn’t pursue electrical engineering degrees, wanting to be like Osokin, because no one knew that he existed. Career advancement required becoming a better bureaucrat, not devising new products or identifying new markets. Civilian products were always an afterthought amid an overwhelming focus on military production.

与此同时,奇怪的是,“复制它”的心态意味着苏联半导体的创新路径是由美国设定的。因此,苏联最敏感和最秘密的行业之一就像硅谷经营不善的前哨一样。Zelenograd 只是全球化网络中的另一个节点——以美国芯片制造商为中心。

Meanwhile, the “copy it” mentality meant, bizarrely, that the pathways of innovation in Soviet semiconductors were set by the United States. One of the most sensitive and secretive industries in the USSR therefore functioned like a poorly run outpost of Silicon Valley. Zelenograd was just another node in a globalizing network—with American chipmakers at the center.

第 9 章 晶体管推销员

CHAPTER 9 The Transistor Salesman

1962 年 11 月,日本首相池田隼人在富丽堂皇的爱丽舍宫会见法国总统戴高乐时,为主人带来了一件小礼物:索尼晶体管收音机。戴高乐是个形式主义和仪式感的人,一个有传统观念的军人,他认为自己是法国威严的化身。相比之下,池田认为他的国家的选民是纯粹的物质主义者,并承诺在十年内将他们的收入翻一番。戴高乐宣称,日本只不过是一个“经济强国”,在会后对一名助手怒吼道:池田表现得像个“晶体管推销员”。但没过多久,全世界都羡慕地注视着日本,因为该国成功销售半导体将使其比戴高乐想象的更富有、更强大。

When Japanese prime minister Hayato Ikeda met French president Charles de Gaulle amid the splendor of the Elysée Palace in November 1962, he brought a small gift for his host: a Sony transistor radio. De Gaulle was formalistic and ceremonious, a tradition-minded military man who saw himself as the incarnation of French grandeur. Ikeda, by contrast, thought his country’s voters were straightforwardly materialistic, and promised to double their incomes within a decade. Japan was nothing but an “economic power,” de Gaulle declared, huffing to an aide after the meeting that Ikeda behaved like a “transistor salesman.” But it wouldn’t be long before all the world was looking enviously at Japan, because the country’s success selling semiconductors would make it far wealthier and more powerful than de Gaulle ever imagined.

集成电路不仅以创新的方式连接电子元件,还以美国为中心将国家连接成一个网络。苏联人通过抄袭硅谷的产品,无意中让自己成为了这个网络的一部分。相比之下,日本有意融入美国的半导体产业,这一过程得到日本商界精英和美国政府的支持。

Integrated circuits didn’t only connect electronic components in innovative ways, they also knit together nations in a network, with the United States at its center. The Soviets inadvertently made themselves part of this network by copying Silicon Valley’s products. Japan, by contrast, was deliberately integrated into America’s semiconductor industry, a process supported by Japanese business elites and the U.S. government.

二战结束时,一些美国人曾设想剥夺日本的高科技产业,以此作为对日本的惩罚。残酷的战争。然而,在日本投降后的几年内,华盛顿的国防官员采取了一项官方政策:“强大的日本比弱小的日本风险更大。” 除了短暂关闭日本对核物理的研究之外,美国政府支持日本重生为科技强国。挑战在于帮助日本重建经济,同时将其与美国主导的体系绑定。让日本成为晶体管推销员是美国冷战战略的核心。

When World War II ended, some Americans had envisioned stripping Japan of its high-tech industries as punishment for starting a brutal war. Yet within a couple years of Japan’s surrender, defense officials in Washington adopted an official policy that “a strong Japan is a better risk than a weak Japan.” Apart from a short-lived effort to shut down Japan’s research into nuclear physics, the U.S. government supported Japan’s rebirth as a technological and scientific power. The challenge was to help Japan rebuild its economy while binding it to an American-led system. Making Japan a transistor salesman was core to America’s Cold War strategy.

关于晶体管发明的消息首先通过统治被占领日本的美国军事当局传播到该国。Makoto Kikuchi 是日本政府位于东京的电工实验室的一名年轻物理学家,该实验室雇用了该国一些最先进的科学家。有一天他的老板把他叫到他的办公室,告诉他一个有趣的消息:菊池的老板解释说,美国科学家已经将两根金属针连接到一个水晶上,并且能够放大电流。菊池知道发现了一个非凡的装置。

News about the invention of the transistor first trickled into the country via the U.S. military authorities who governed occupied Japan. Makoto Kikuchi was a young physicist in the Japanese government’s Electrotechnical Laboratory in Tokyo, which employed some of the country’s most advanced scientists. One day his boss called him into his office with interesting news: American scientists, Kikuchi’s boss explained, had attached two metal needles to a crystal and were able to amplify a current. Kikuchi knew an extraordinary device had been discovered.

在被炸毁的东京,人们很容易感到与世界领先的物理学家隔绝,但东京的美国占领总部为日本科学家提供了访问贝尔系统技术杂志应用物理学杂志物理评论等期刊的机会,这些期刊发表了巴丁、布拉顿和肖克利。这些期刊在战后的日本是不可能获得的。“我会浏览内容,每当我看到‘半导体’或‘晶体管’这个词时,”菊池回忆道,“我的心会开始跳动。” 几年后,在 1953 年,当这位美国科学家在炎热潮湿的 9 月前往东京参加国际纯粹与应用物理学联合会会议时,菊池遇到了约翰·巴丁。Bardeen 被视为名人,他对想要为他拍照的人数感到震惊。“我这辈子从没见过这么多闪光灯,”他写信给妻子。

In bombed-out Tokyo, it was easy to feel isolated from the world’s leading physicists, but U.S. occupation headquarters in Tokyo provided Japan’s scientists access to journals like Bell System Technical Journal, Journal of Applied Physics, and Physical Review, which published the papers of Bardeen, Brattain, and Shockley. These journals were otherwise impossible to obtain in postwar Japan. “I’d flick through the contents and whenever I saw the word ‘semiconductor’ or ‘transistor,’ ” Kikuchi recounted, “my heart would start to pound.” Several years later, in 1953, Kikuchi met John Bardeen when the American scientist traveled to Tokyo during a hot and humid September for a meeting of the International Union of Pure and Applied Physics. Bardeen was treated like a celebrity and was shocked by the number of people wanting to take his photo. “I’ve never seen so many flashbulbs in my life,” he wrote his wife.

同年巴丁登陆东京,森田昭夫从羽田机场起飞前往纽约。作为日本最杰出的清酒酿酒厂之一的第十五代继承人,森田已经得到了培养从出生就接手家族企业。森田的父亲曾希望儿子成为第十六代经营清酒业务的森田,但森田昭夫从小就喜欢修补电子学和大学物理学学位,却指向了不同的方向。在战争期间,这种物理专业知识可能挽救了他的生命,让他被送到研究实验室而不是前线。

The same year Bardeen landed in Tokyo, Akio Morita took off from Haneda Airport for New York. The fifteenth-generation heir to one of Japan’s most distinguished sake distilleries, Morita had been groomed since birth to take over the family business. Morita’s father had wanted his son to become the sixteenth Morita to manage the sake business, but Akio Morita’s childhood love of tinkering with electronics and a university degree in physics pointed in a different direction. During the war, this physics expertise may have saved his life, getting him sent to a research lab rather than the front lines.

森田的物理学学位在战后的日本也被证明是有用的。1946 年 4 月,在这个国家仍处于废墟之中的情况下,森田与一位名叫 Masaru Ibuka 的前同事合作建立了一家电子业务,他们很快将其命名为 Sony,取自拉丁语sonus(声音)和美国昵称“sonny”。他们的第一个设备,电饭煲,是个废品,但他们的录音机工作得很好,卖得更好。1948 年,森田读到了贝尔实验室的新晶体管并立即抓住了它的潜力。森田回忆说,这似乎是“奇迹”,梦想着彻底改变消费设备。

Morita’s physics degree proved useful in postwar Japan, too. In April 1946, with the country still in ruins, Morita partnered with a former colleague named Masaru Ibuka to build an electronics business, which they soon named Sony, from the Latin sonus (sound) and the American nickname “sonny.” Their first device, an electric rice cooker, was a dud, but their tape recorder worked well and sold better. In 1948, Morita read about Bell Labs’ new transistor and immediately grasped its potential. It seemed “miraculous,” Morita recalled, dreaming of revolutionizing consumer devices.

1953 年登陆美国后,森田对这个国家广阔的距离、开放的空间和非凡的消费财富感到震惊,尤其是与战后东京的剥夺相比。这个国家似乎什么都有,森田想。在纽约,他遇到了 AT&T 高管,他们同意向他颁发生产晶体管的许可证。他们告诉他不要期望制造比助听器更有用的东西。

Upon landing in the United States in 1953, Morita was shocked by the country’s vast distances, open spaces, and extraordinary consumer wealth, especially compared to the deprivation of postwar Tokyo. This country seems to have everything, Morita thought. In New York, he met AT&T executives who agreed to issue him a license to produce the transistor. They told him to expect to manufacture nothing more useful than a hearing aid.

森田明白查尔斯·戴高乐不明白的事情:电子产品是世界经济的未来,而很快嵌入硅芯片中的晶体管将使难以想象的新设备成为可能。森田意识到,晶体管提供的更小尺寸和更低功耗将改变消费电子产品。他和 Ibuka 决定将他们公司的未来赌在不仅向日本客户销售这些设备上,而且向世界上最富有的消费市场美国销售这些设备。

Morita understood what Charles de Gaulle did not: electronics were the future of the world economy, and transistors, soon embedded in silicon chips, would make possible unimaginable new devices. The smaller size and lower power consumption that transistors offered, Morita realized, were set to transform consumer electronics. He and Ibuka decided to bet the future of their company on selling these devices not only to Japanese customers, but to the world’s richest consumer market, America.

日本政府表示支持高科技,同年,日本皇太子访问了美国无线电研究实验室,森田前往贝尔实验室。日本强大的国际部贸易和工业部也希望支持电子公司,但该部的影响喜忧参半,官僚们一度将索尼从贝尔实验室获得晶体管许可的申请推迟了几个月,理由是它是该公司未经该部同意就与一家外国公司签订合同,这“不可原谅”。

Japan’s government signaled its support for high tech, with Japan’s crown prince visiting an American radio research lab the same year Morita traveled to Bell Labs. Japan’s powerful Ministry of International Trade and Industry also wanted to support electronics firms, but the ministry’s impact was mixed, with bureaucrats at one point delaying Sony’s application to license the transistor from Bell Labs by several months on the grounds that it was “inexcusably outrageous” for the company to have signed a contract with a foreign firm without the ministry’s consent.

索尼在日本获得了较低工资的好处,但它的商业模式归根结底是关于创新、产品设计和营销的。森田的“许可”策略与苏联部长肖金的“复制”策略大相径庭。许多日本公司以无情的制造效率而闻名。索尼擅长发现新市场,并使用硅谷最新的电路技术为他们提供令人印象深刻的产品。“我们的计划是用新产品引导公众,而不是问他们想要什么样的产品,”森田宣称。“公众不知道什么是可能的,但我们知道。”

Sony had the benefit of cheaper wages in Japan, but its business model was ultimately about innovation, product design, and marketing. Morita’s “license it” strategy couldn’t have been more different from the “copy it” tactics of Soviet Minister Shokin. Many Japanese companies had reputations for ruthless manufacturing efficiency. Sony excelled by identifying new markets and targeting them with impressive products using Silicon Valley’s newest circuitry technology. “Our plan is to lead the public with new products rather than ask them what kind of products they want,” Morita declared. “The public does not know what is possible, but we do.”

索尼的第一个重大成功是晶体管收音机,比如池田首相送给戴高乐的收音机。几年前,德州仪器曾试图推销晶体管收音机,但尽管它拥有必要的技术,但德州仪器在定价和营销方面搞砸了,并很快放弃了这项业务。森田看到了一个机会,很快就生产出数以万计的设备。

Sony’s first major success was transistor radios, such as the radio Prime Minister Ikeda had given de Gaulle. Several years earlier, Texas Instruments had tried to market transistor radios, but though it had the necessary technology, TI bungled the pricing and marketing and quickly abandoned the business. Morita saw an opening and was soon churning out tens of thousands of the devices.

尽管如此,像仙童这样的美国芯片公司继续在芯片生产的前沿占据主导地位,例如其与企业大型计算机相关的业务。在整个 1960 年代,日本公司为知识产权支付了可观的许可费,将所有芯片销售额的 4.5% 交给飞兆半导体,将 3.5% 交给德州仪器,将 2% 交给西部电气。美国芯片制造商乐于转让他们的技术,因为日本公司似乎落后了很多年。

Nevertheless, U.S. chip firms like Fairchild continued to dominate the cutting edge of chip production, such as its business related to corporate mainframe computers. Throughout the 1960s, Japanese firms paid sizeable licensing fees on intellectual property, handing over 4.5 percent of all chip sales to Fairchild, 3.5 percent to Texas Instruments, and 2 percent to Western Electric. U.S. chipmakers were happy to transfer their technology because Japanese firms appeared to be years behind.

索尼的专长不是设计芯片,而是设计消费产品和定制他们需要的电子产品。计算器是日本公司改造的另一种消费设备。拍1967 年,TI 主席 Haggerty 曾要求 Jack Kilby 制造一款手持式、半导体驱动的计算器。然而,TI 的营销部门认为廉价的手持式计算器没有市场,因此该项目停滞不前。日本的夏普电子不同意,将加利福尼亚生产的芯片放入计算器中,这比任何人想象的都要简单和便宜。夏普的成功保证了 1970 年代生产的大多数计算器都是日本制造的。要是 TI 能早点找到销售自有品牌设备的方法就好了,Haggerty 后来感叹道,TI“将成为消费电子领域的索尼”。然而,复制索尼的产品创新和营销专长与复制美国的半导体专长一样困难。

Sony’s expertise wasn’t in designing chips but devising consumer products and customizing the electronics they needed. Calculators were another consumer device transformed by Japanese firms. Pat Haggerty, the TI Chairman, had asked Jack Kilby to build a handheld, semiconductor-powered calculator in 1967. However, TI’s marketing department didn’t think there’d be a market for a cheap, handheld calculator, so the project stagnated. Japan’s Sharp Electronics disagreed, putting California-produced chips in a calculator that was far simpler and cheaper than anyone had thought possible. Sharp’s success guaranteed most calculators produced in the 1970s were Japanese made. If only TI had found a way to market its own branded devices earlier, Haggerty later lamented, TI “would have been the Sony of consumer electronics.” Replicating Sony’s product innovation and marketing expertise, however, proved just as hard as replicating America’s semiconductor expertise.

美国和日本之间出现的半导体共生关系涉及复杂的平衡行为。每个国家都依赖对方的供应和客户。到 1964 年,日本的分立晶体管产量已超过美国,而美国公司则生产最先进的芯片。美国公司制造了最好的电脑,而索尼和夏普等电子产品制造商则生产推动半导体消费的消费品。日本的电子产品出口——包括半导体和依赖它们的产品——从 1965 年的 6 亿美元激增至大约 20 年后的 600 亿美元。

The semiconductor symbiosis that emerged between America and Japan involved a complex balancing act. Each country relied on the other for supplies and for customers. By 1964, Japan had overtaken the U.S. in production of discrete transistors, while American firms produced the most advanced chips. U.S. firms built the best computers, while electronics manufacturers like Sony and Sharp produced consumer goods that drove semiconductor consumption. Japan’s exports of electronics—a mix of semiconductors and products that relied on them—boomed from $600 million in 1965 to $60 billion around two decades later.

相互依赖并不总是那么容易。1959年,电子工业协会呼吁美国政府提供帮助,以免日本进口破坏“国家安全”——以及他们自己的底线。但让日本建立电子产业是美国冷战战略的一部分,因此,在 1960 年代,华盛顿从未在这个问题上对东京施加太大压力。像《电子杂志》这样的贸易刊物——原本可能会站在美国公司一边——反而指出,“日本是美国太平洋政策的基石…… 如果她不能与西半球和欧洲进行健康的商业往来,她将在其他地方寻求经济支持,”比如共产主义中国或苏联。我们战略要求让日本获得先进技术并建立尖端业务。“一个有他们历史的人不会满足于制造晶体管收音机,”理查德尼克松总统后来评论道。必须允许,甚至鼓励他们开发更先进的技术。

Interdependence wasn’t always easy. In 1959, the Electronics Industries Association appealed to the U.S. government for help lest Japanese imports undermine “national security”—and their own bottom line. But letting Japan build an electronics industry was part of U.S. Cold War strategy, so, during the 1960s, Washington never put much pressure on Tokyo over the issue. Trade publications like Electronics magazine—which might have been expected to take the side of U.S. companies—instead noted that “Japan is a keystone in America’s Pacific policy…. If she cannot enter into healthy commercial intercourse with the Western hemisphere and Europe, she will seek economic sustenance elsewhere,” like Communist China or the Soviet Union. U.S. strategy required letting Japan acquire advanced technology and build cutting-edge businesses. “A people with their history won’t be content to make transistor radios,” President Richard Nixon later observed. They had to be allowed, even encouraged, to develop more advanced technology.

日本高管同样致力于使这种半导体共生发挥作用。当德州仪器试图成为第一家在日本开设工厂的外国芯片制造商时,该公司面临着重重的监管障碍。索尼的森田恰好是哈格蒂的朋友,他提出提供帮助以换取一部分利润。他告诉 TI 高管隐身访问东京,以假名在他们的酒店注册,并且永远不要离开他们的酒店房间。森田偷偷造访酒店,提议成立合资公司:TI在日本生产芯片,索尼管理官僚。“我们将为您提供保障,”他告诉德州仪器公司的高管。德州人认为索尼是“流氓行动”,他们的意思是恭维。

Japanese executives were no less committed to making this semiconductor symbiosis work. When Texas Instruments sought to become the first foreign chipmaker to open a plant in Japan, the company faced a thicket of regulatory barriers. Sony’s Morita, who happened to be a friend of Haggerty, offered to help in exchange for a share of the profits. He told TI executives to visit Tokyo incognito, register at their hotel under false names, and never leave their hotel room. Morita visited the hotel clandestinely and proposed a joint venture: TI would produce chips in Japan, and Sony would manage the bureaucrats. “We will cover for you,” he told the Texas Instruments executives. The Texans thought Sony was a “rogue operation,” something they meant as a compliment.

在森田的帮助下,经过大量繁文缛节和绿茶,日本官僚最终批准了 TI 在日本开设半导体工厂的许可。对森田来说,这又是一次成功,使他成为太平洋两岸最著名的日本商人之一。对于华盛顿的外交政策战略家来说,两国之间更多的贸易和投资联系将东京与美国主导的体系联系得更加紧密。对于像池田首相这样的日本领导人来说,这也是一场胜利。他让日本人的收入翻倍的目标在两年内就实现了提前于计划。多亏像森田这样勇敢的电子企业家,日本在世界舞台上赢得了新的席位。晶体管推销员的影响力远比戴高乐想象的要大得多。

With Morita’s help, and after much red tape and green tea, Japan’s bureaucrats finally approved TI’s permits to open a semiconductor plant in Japan. For Morita, it was another coup, helping to make him one of the most famous Japanese businessmen on either side of the Pacific. For foreign policy strategists in Washington, more trade and investment links between the two countries tied Tokyo ever more tightly into a U.S.-led system. It was a victory for Japanese leaders like Prime Minister Ikeda, too. His goal of doubling Japanese incomes was achieved two years ahead of schedule. Japan won a new seat on the world stage thanks to intrepid electronics entrepreneurs like Morita. Transistor salesman was a position of far more influence than Charles de Gaulle could ever have imagined.

第 10 章 “晶体管女孩”

CHAPTER 10 “Transistor Girls”

他们的衣服是西方的,但他们的爱情仪式是建立在东方古老的快乐之上的,”封面上写道《晶体管女孩》,一部 1964 年的垃圾澳大利亚小说。情节涉及中国黑帮、国际阴谋和“通过课外活动增加收入”的流水线女工。《晶体管女孩》封面上的图片显示了一名年轻的日本女性,衣着暴露,背景是一座宝塔的轮廓。封底展示了一位女性,她身在东方主义意象中,但衣着更少。

“Their clothing was of the West, but their love rites were founded in the ancient pleasures of the East,” read the cover of The Transistor Girls, a trashy Australian novel from 1964. The plot involved Chinese gangsters, international intrigue, and women assembly line workers who “added to their incomes by extracurricular night-time activity.” The image on the front cover of The Transistor Girls showed a young Japanese woman, scantily clad, with a silhouette of a pagoda in the background. The back cover revealed a woman amid more orientalist imagery but with even less clothing.

设计最早的半导体的主要是男性,组装它们的主要是女性。摩尔定律预测计算能力的成本即将暴跌。但是,让摩尔的愿景成为现实不仅仅是缩小芯片上每个晶体管尺寸的问题。它还需要更多、更便宜的工人来组装它们。

It was mostly men who designed the earliest semiconductors, and mostly women who assembled them. Moore’s Law predicted the cost of computing power was about to plummet. But making Moore’s vision a reality wasn’t only a question of shrinking the size of each transistor on a chip. It also required a larger and cheaper supply of workers to assemble them.

Fairchild Semiconductor 的许多员工加入公司是为了寻找财富或因为对工程的热爱。Charlie Sporck 在被赶出之前的工作后来到 Fairchild。一个抽着雪茄,努力驾驶的纽约人,Sporck 专注于效率。在一个充满才华横溢的科学家和技术远见者的行业中,Sporck 的专长在于提高工人的生产力和机器一样。多亏了像他这样强硬的管理者,计算成本才符合戈登摩尔预测的进度。

Many employees of Fairchild Semiconductor joined the firm in search of riches or because of a love of engineering. Charlie Sporck came to Fairchild after being chased out of his previous job. A cigar smoking, hard-driving New Yorker, Sporck was fixated on efficiency. In an industry full of brilliant scientists and technological visionaries, Sporck’s expertise was in wringing productivity out of workers and machines alike. It was only thanks to tough managers like him that the cost of computing fell in line with the schedule Gordon Moore had predicted.

Sporck 在 1950 年代中期被通用电气公司聘用在纽约哈德逊瀑布的工厂之前,曾在康奈尔大学学习工程学。他的任务是改进 GE 制造电容器的流程,并提议改变工厂的装配线流程。他相信他的新技术会提高生产力,但控制通用电气装配线工人的工会认为他威胁到他们对生产过程的控制。工会起义,集会反对斯波克并将他的肖像烧死。工厂的管理层胆怯地退缩了,向工会承诺,Sporck 的改变永远不会实施。

Sporck had studied engineering at Cornell before being hired by GE in the mid-1950s at the firm’s factory in Hudson Falls, New York. He was tasked with improving GE’s process for manufacturing capacitors and proposed changing the factory’s assembly line process. He believed his new technique would improve productivity, but the labor union that controlled GE’s assembly line workers saw him as threatening their control over the production process. The union revolted, staging a rally against Sporck and burning him in effigy. The factory’s management timidly backed down, promising the union that Sporck’s changes would never be implemented.

见鬼去吧,斯波克想。那天晚上,他回到家,开始寻找其他工作。1959 年 8 月,他在《华尔街日报》上看到了一则广告,要求在一家名为仙童半导体的小公司担任生产经理职位,并提交了一份申请。不久,他被叫到纽约市列克星敦大道的一家酒店接受采访。采访他的两名 Fairchild 员工在午餐后喝得酩酊大醉,当场给了他一份工作。这是 Fairchild 做出的最佳招聘决定之一。斯波克从未在俄亥俄州以西,但他立即接受了,此后不久就到山景城报到。

To hell with this, Sporck thought. That night, he arrived home and started looking for other jobs. In August 1959, he saw an ad in the Wall Street Journal for a production manager role at a small company called Fairchild Semiconductor and sent in an application. Soon he was called into New York City for an interview in a hotel on Lexington Avenue. The two Fairchild employees who interviewed him were drunk after a boozy lunch and offered him a job on the spot. It was one of the best hiring decisions Fairchild made. Sporck had never been west of Ohio, but he accepted immediately, reporting for duty in Mountain View shortly thereafter.

斯波克回忆说,抵达加利福尼亚后,他惊讶地发现该公司“在处理劳工和工会方面几乎没有能力。我把这种能力带给了我的新雇主。” 许多公司不会将最终导致管理层被烧毁的劳资关系战略描述为“有能力”。但在硅谷,工会很弱,而斯波克致力于保持这种状态。他宣称,他和他在 Fairchild 的同事对工会“死心塌地”。作为一名务实、脚踏实地的工程师,Sporck 并不是一个刻板的工会破坏者。他的办公室如此简朴,以至于被比作军营。斯波克很自豪为大多数员工提供股票期权,这种做法在东海岸的老电子公司中几乎是不为人知的。但他会无情地坚持,作为交换,这些员工承诺最大化他们的生产力。

Upon arrival in California, Sporck recalled, he was surprised that the firm “had virtually no competence in the handling of labor and labor unions. I brought this competence to my new employer.” Many companies wouldn’t have described a strategy of labor relations that culminated in management getting burned in effigy as “competent.” But in Silicon Valley, unions were weak, and Sporck was committed to keeping it that way. He and his colleagues at Fairchild were “dead set” against unions, he declared. A practical, down-to-earth engineer, Sporck wasn’t a stereotypical union buster. He kept his offices so austere that they were compared to an army barracks. Sporck was proud of giving most employees stock options, a practice that was virtually unknown in the old East Coast electronics firms. But he’d ruthlessly insist, in exchange, that these same employees commit to maximizing their productivity.

与劳动力往往以男性为主的东海岸电子公司不同,旧金山以南的大多数新芯片初创公司他们的装配线配备了女性。几十年来,妇女一直在圣克拉拉谷从事流水线工作,首先是在 1920 年代和 1930 年代推动该谷经济发展的水果罐头厂,然后是二战期间的航空航天业。国会 1965 年放宽移民规则的决定使许多外国出生的女性加入了山谷的劳动力库。

Unlike East Coast electronics firms whose workforces tended to be male-dominated, most of the new chip startups south of San Francisco staffed their assembly lines with women. Women had worked in assembly line jobs in the Santa Clara Valley for decades, first in the fruit canneries that drove the valley’s economy in the 1920s and 1930s, then in the aerospace industry during World War II. Congress’s decision to ease immigration rules in 1965 added many foreign-born women to the valley’s labor pool.

芯片公司雇用女性是因为她们可以得到较低的工资,而且与男性相比,她们不太可能要求更好的工作条件。生产经理还认为,女性的小手使她们更擅长组装和测试成品半导体。在 1960 年代,将硅芯片连接到塑料片上的过程首先需要通过显微镜观察,以将硅片定位到塑料上。然后,组装工人将这两部分固定在一起,同时机器施加热量、压力和超声波振动,将硅粘合到塑料底座上。再次用手连接细金线,以在芯片之间进行导电。最后,必须将芯片插入仪表进行测试——另一个当时只能手工完成的步骤。随着对芯片的需求猛增,对可以组装它们的双手的需求也在飙升。

Chip firms hired women because they could be paid lower wages and were less likely than men to demand better working conditions. Production managers also believed women’s smaller hands made them better at assembling and testing finished semiconductors. In the 1960s, the process of attaching a silicon chip to the piece of plastic on which it would sit first required looking through a microscope to position the silicon onto the plastic. The assembly worker then held the two pieces together as a machine applied heat, pressure, and ultrasonic vibration to bond the silicon to the plastic base. Thin gold wires were attached, again by hand, to conduct electricity to and from the chip. Finally, the chip had to be tested by plugging it into a meter—another step that at the time could only be done by hand. As demand for chips skyrocketed, so did the demand for pairs of hands that could assemble them.

无论他们在加利福尼亚的哪个地方看,像斯波克这样的半导体高管都找不到足够的廉价工人。Fairchild 遍及美国,最终在缅因州开设了工厂——那里的工人“憎恨工会”,Sporck 报道——以及在新墨西哥州的一个提供税收优惠的纳瓦霍保留地。然而,即使在美国最贫穷的地区,劳动力成本也很高。鲍勃·诺伊斯(Bob Noyce)在香港的一家无线电组装厂进行了个人投资,香港是与毛泽东的共产党人接壤的英国殖民地。中国。工资是美国平均水平的十分之一——大约每小时 25 美分。“你为什么不去看看,”诺伊斯告诉斯波克,他是很快就上飞机去看看。

Wherever they looked across California, semiconductor executives like Sporck couldn’t find enough cheap workers. Fairchild scoured the U.S., eventually opening facilities in Maine—where workers had “a hatred for the labor unions,” Sporck reported—and on a Navajo reservation in New Mexico that provided tax incentives. Even in the poorest parts of America, however, labor costs were substantial. Bob Noyce had made a personal investment in a radio assembly factory in Hong Kong, the British colony just across the border from Mao Zedong’s Communist China. Wages were a tenth of the American average—around 25 cents an hour. “Why don’t you go take a look,” Noyce told Sporck, who was soon on a plane to check it out.

Fairchild 的一些同事很担心。“红色中国人在你的鼻子上,”一个人警告说,注视着驻扎在香港北部边境的数千名中国人民解放军士兵。“你会被撞倒的。” 但诺伊斯投资的收音机工厂说明了这个机会。“中国劳工,那里工作的女孩,超出了我们所知道的一切,”Sporck 的一位同事回忆道。Fairchild 高管认为,香港的装配工人速度似乎是美国人的两倍,等等“愿意忍受单调的工作,”一位高管报告说。

Some colleagues at Fairchild were apprehensive. “The Red Chinese are down your nose,” one warned, eying the thousands of People’s Liberation Army soldiers stationed on Hong Kong’s northern border. “You’re going to get run over.” But the radio factory Noyce had invested in illustrated the opportunity. “The Chinese labor, the girls working there, were exceeding everything that was ever known,” one of Sporck’s colleagues recalled. Assembly workers in Hong Kong seemed twice as fast as Americans, Fairchild executives thought, and more “willing to tolerate monotonous work,” one executive reported.

仙童在恒业街的一家凉鞋厂租用了场地,毗邻旧香港机场,就在九龙湾岸边。很快,几层楼高的巨大飞兆标志就安装在大楼上,照亮了在香港港口周围航行的帆船。Fairchild 继续在加利福尼亚生产其硅晶片,但开始将半导体运送到香港进行最终组装。1963 年,香港工厂投入运营的第一年,组装了 1.2 亿台设备。生产质量非常好,因为低劳动力成本意味着 Fairchild 可以聘请训练有素的工程师来运行装配线,而这本来是在加州贵得让人望而却步。

Fairchild rented space in a sandal factory on Hang Yip Street, next to the old Hong Kong airport, right on the shore of Kowloon Bay. Soon a huge Fairchild logo several stories tall was mounted on the building, illuminating the junks sailing around Hong Kong’s harbor. Fairchild continued to make its silicon wafers in California but began shipping semiconductors to Hong Kong for final assembly. In 1963, its first year of operation, the Hong Kong facility assembled 120 million devices. Production quality was excellent, because low labor costs meant Fairchild could hire trained engineers to run assembly lines, which would have been prohibitively expensive in California.

Fairchild 是第一家在亚洲进行离岸组装的半导体公司,但德州仪器、摩托罗拉和其他公司很快跟进。十年之内,几乎所有美国芯片制造商都拥有外国组装设施。Sporck 开始将目光投向香港以外的地方。该市 25 美分的时薪仅为美国工资的十分之一,但在亚洲名列前茅。在 1960 年代中期,台湾工人每小时赚 19 美分,马来西亚人 15 美分,新加坡人 11 美分,韩国人只有一角钱。

Fairchild was the first semiconductor firm to offshore assembly in Asia, but Texas Instruments, Motorola, and others quickly followed. Within a decade, almost all U.S. chipmakers had foreign assembly facilities. Sporck began looking beyond Hong Kong. The city’s 25-cent hourly wages were only a tenth of American wages but were among the highest in Asia. In the mid-1960s, Taiwanese workers made 19 cents an hour, Malaysians 15 cents, Singaporeans 11 cents, and South Koreans only a dime.

Sporck 的下一站是新加坡,这是一个以华人为主的城邦,其领导人李光耀曾正如一位仙童老兵所记得的那样,“几乎是非法的”工会。仙童紧随其后开幕不久之后在马来西亚槟城市的一个设施。半导体行业在人们听说这个词之前几十年就已经全球化,为我们今天所知的以亚洲为中心的供应链奠定了基础。

Sporck’s next stop was Singapore, a majority ethnic Chinese city-state whose leader, Lee Kuan Yew, had “pretty much outlawed” unions, as one Fairchild veteran remembered. Fairchild followed by opening a facility in the Malaysian city of Penang shortly thereafter. The semiconductor industry was globalizing decades before anyone had heard of the word, laying the grounds for the Asia-centric supply chains we know today.

像 Sporck 这样的经理人没有全球化的游戏计划。如果成本相同,他会很高兴继续在缅因州或加利福尼亚州建造工厂。但是亚洲有数以百万计的农民在寻找工厂工作,保持低工资并保证他们会在一段时间内保持低水平。华盛顿的外交政策战略家认为,香港、新加坡和槟城等城市的华裔工人已经成熟,可以接受毛泽东的共产党颠覆。Sporck 将它们视为资本家的梦想。“我们在硅谷遇到了工会问题,”斯波克指出。“我们在东方从来没有遇到过任何工会问题。”

Managers like Sporck had no game plan for globalization. He’d just as happily have kept building factories in Maine or California had they cost the same. But Asia had millions of peasant farmers looking for factory jobs, keeping wages low and guaranteeing they’d stay low for some time. Foreign policy strategists in Washington saw ethnic Chinese workers in cities like Hong Kong, Singapore, and Penang as ripe for Mao Zedong’s Communist subversion. Sporck saw them as a capitalist’s dream. “We had union problems in Silicon Valley,” Sporck noted. “We never had any union problems in the Orient.”

第 11 章 精准打击

CHAPTER 11 Precision Strike

回合1970 年代初期,在公司位于新加坡和香港的半导体工厂之间的飞行途中,德州仪器 (TI) 的员工偶尔会凝视飞机窗外,俯视越南沿海平原战场上升起的烟雾。TI 在亚洲的员工专注于制造芯片,而不是战争。然而,他们在得克萨斯州的许多同事却一无所获。TI 的第一个主要集成电路合同是为 Minuteman II 等大型核导弹签订的,但越南战争需要不同类型的武器。越南早期的轰炸行动,例如从 1965 年到 1968 年的“滚雷行动”,都失败了八十万吨炸弹,比二战期间在太平洋战区投掷的炸弹还多。然而,这种火力对北越军队的影响很小,因为大多数炸弹都没有击中目标。

About halfway on the flight between the company’s semiconductor plants in Singapore and Hong Kong during the early 1970s, Texas Instruments employees would occasionally peer out their aircraft window and look down on puffs of smoke rising from the battlefields on Vietnam’s coastal plains. TI’s staff across Asia were focused on making chips, not on the war. Many of their colleagues in Texas, however, thought about nothing else. TI’s first major contract for integrated circuits had been for massive nuclear missiles like the Minuteman II, but the war in Vietnam required different types of weapons. The early bombing campaigns in Vietnam, like Operation Rolling Thunder, which stretched from 1965 to 1968, dropped over eight hundred thousand tons of bombs, more than was dropped in the Pacific Theater during all of World War II. This firepower had only a marginal impact on North Vietnam’s military, however, because most of the bombs missed their targets.

空军意识到它需要更聪明地战斗。军方已经尝试了多种技术来引导其导弹和炸弹,从使用遥控器到红外导引头。其中一些武器,例如伯劳导弹,它从飞机上发射并使用简单的制导系统将导弹指向雷达无线电波的源头并瞄准敌方雷达设施,证明合理有效。但许多其他制导系统似乎几乎无法工作。直到 1985 年,国防部的一项研究发现只有四个空对空导弹在视距外击落敌机的例子。有了这样的限制,制导弹药似乎不可能决定战争的结果。

The Air Force realized it needed to fight smarter. The military had experimented with a variety of techniques for guiding its missiles and bombs, from using remote control to infrared seekers. Some of these weapons, like the Shrike missile, which was launched from planes and homed in on enemy radar facilities using a simple guidance system that pointed the missile toward the source of the radar’s radio waves, proved reasonably effective. But many other guidance systems seemed hardly ever to work. As late as 1985, a Defense Department study found only four examples of air-to-air missiles downing an enemy aircraft outside of visual range. With limitations like these, it seemed impossible that guided munitions would ever decide the outcome of a war.

军方总结说,许多制导弹药的问题在于真空管。美国战斗机在越南上空使用的麻雀 III 防空导弹依赖于手工焊接的真空管。东南亚潮湿的气候,起降的力量,战机战斗的颠簸,经常导致失败。麻雀导弹的雷达系统平均每使用五到十小时就会损坏一次。战后的一项研究发现,在越南发射的麻雀只有 9.2% 击中了目标,而 66% 发生故障,并且其余的只是错过了。

The problem with many guided munitions, the military concluded, was the vacuum tubes. The Sparrow III anti-aircraft missiles that U.S. fighters used in the skies over Vietnam relied on vacuum tubes that were hand-soldered. The humid climate of Southeast Asia, the force of takeoff and landings, and the rough-and-tumble of fighter combat caused regular failures. The Sparrow missile’s radar system broke on average once every five to ten hours of use. A postwar study found that only 9.2 percent of Sparrows fired in Vietnam hit their target, while 66 percent malfunctioned, and the rest simply missed.

然而,军方在越南面临的最大挑战是打击地面目标。在越南战争开始时,炸弹平均掉落根据空军数据,距离目标 420 英尺以内。因此,用炸弹攻击车辆基本上是不可能的。34 岁的 TI 项目工程师 Weldon Word 想要改变这种状况。Word 有一双透彻的蓝眼睛,响亮、深沉、催眠的声音,以及思考战争未来的独特优势。他刚刚结束了在一艘海军舰艇上为 TI 开发的新声纳收集数据的为期一年的工作,这项任务单调得让人麻木,但这表明军事系统可以通过正确的传感器和仪器收集多少数据。早在 1960 年代中期,Word 就已经在设想使用微电子技术来改变军队的杀伤链。卫星和飞机上的先进传感器将捕获目标、跟踪目标、引导导弹飞向目标,并确认它们已被销毁。这听起来像科幻小说。但TI 已经在其研究实验室中生产了必要的组件。

The military’s biggest challenge in Vietnam, however, was striking ground targets. At the start of the Vietnam War, bombs fell on average within 420 feet of their target, according to Air Force data. Attacking a vehicle with a bomb was therefore basically impossible. Weldon Word, a thirty-four-year-old project engineer at TI, wanted to change this. Word had penetrating blue eyes, a loud, deep, hypnotic voice, and a unique vantage point for thinking about the future of war. He’d just concluded a yearlong stint aboard a Navy ship gathering data for a new TI-developed sonar, an assignment that was mind-numbingly monotonous, but that demonstrated how much data military systems could collect with the right sensors and instrumentation. As early as the mid-1960s, Word was already envisioning using microelectronics to transform the military’s kill chain. Advanced sensors on satellites and in airplanes would acquire targets, track them, guide missiles toward them, and confirm they were destroyed. It sounded like science fiction. But TI already produced the necessary components in its research labs.

TI 为其制造芯片的洲际弹道导弹提出了相对简单的制导挑战。他们是从地面上的固定位置发射,而不是从以每小时数百英里的速度飞行的飞机发射,同时机动以避免敌人的火力。洲际弹道导弹目标也没有移动。导弹本身仅受到风和天气条件的轻微影响,因为它们以数倍音速从外太空向下倾斜。他们携带的弹头足够大,即使是轻微的失误也会造成极大的破坏。从蒙大拿州攻击莫斯科比用一架 F-4 投下的炸弹击中一辆卡车要容易得多,而 F-4 在几千英尺的高度飞行。

The intercontinental ballistic missiles that TI had built chips for presented a relatively straightforward guidance challenge. They were launched from a fixed position on the ground, not from a plane flying at several hundred miles per hour while maneuvering to avoid enemy fire. ICBM targets didn’t move either. The missiles themselves were only slightly impacted by wind and weather conditions as they careened downward from outer space at multiple times the speed of sound. They carried warheads big enough to make even a slight miss immensely destructive. It was vastly easier to hit Moscow from Montana than it was to hit a truck with a bomb dropped by an F-4 flying at a couple thousand feet.

这是一项复杂的任务,但 Word 明白最好的武器是“便宜又熟悉,”他的一位同事解释说,保证它们可以经常在训练和战场上使用。微电子必须设计得尽可能简单。每个必须焊接的连接都会增加可靠性风险。电子设备越简单,系统就越可靠和节能。

This was a complex task, but Word understood that the best weapons were “cheap and familiar,” one of his colleagues explained, guaranteeing that they could be used often in training and on the battlefield. The microelectronics had to be designed with as little complexity as possible. Every connection that had to be soldered increased the risks to reliability. The simpler the electronics, the more reliable and more power-efficient a system would be.

许多国防承包商试图向五角大楼出售昂贵的导弹,但 Word 告诉他的团队制造武器价格像便宜的家用轿车。他一直在寻找一种简单易用的设备,使其能够快速部署在各种类型的飞机上,被每个军种所接受,并迅速被美国盟友采用。

Many defense contractors were trying to sell the Pentagon expensive missiles, but Word told his team to build weapons priced like an inexpensive family sedan. He was on the lookout for a device that was simple and easy to use, enabling it to be quickly deployed on every type of airplane, embraced by each military service, and quickly adopted by U.S. allies, too.

1965 年 6 月,Word 飞往佛罗里达州的埃格林空军基地,在那里他遇到了负责采购新设备以供越南使用的计划的军官乔·戴维斯上校。戴维斯在 15 岁时就学会了飞行,然后就加入了军队,并在二战和韩国执行过战斗机和轰炸机任务。之后,他指挥了欧洲和太平洋的空军部队。他比任何人都清楚什么样的武器可以在空军任务中发挥作用。当 Word 在他的办公室坐下时,戴维斯打开他的办公桌抽屉,拿出一张清化桥的照片,这是一座 540 英尺长的金属结构,横跨北越松马河,四周环绕着空气防御。Word 和 Davis 数了数桥周围有 800 个麻点,每个麻点都是由未命中目标的美国炸弹或火箭造成的。几十个,也许还有几百个炸弹掉进了河里,没有留下任何痕迹。桥依然屹立不倒。德州仪器能否提供帮助?戴维斯问道。

In June 1965, Word flew to Florida’s Eglin Air Force Base, where he met Colonel Joe Davis, the officer in charge of a program to acquire new equipment for use in Vietnam. Davis had learned to fly at age fifteen before joining the military and piloting both fighter and bomber missions in World War II and Korea. Afterward he commanded Air Force units both in Europe and the Pacific. He understood better than anyone what type of weapons would work in Air Force missions. When Word sat down in his office, Davis opened his desk drawer and pulled out a photo of the Thanh Hoa Bridge, a 540-foot-long metallic structure stretching across North Vietnam’s Song Ma river, ringed with air defenses. Word and Davis counted eight hundred pockmarks around the bridge, each caused by an American bomb or rocket that missed its target. Dozens, maybe hundreds more bombs had fallen in the river and left no mark. The bridge was still standing. Could Texas Instruments do anything to help? Davis asked.

Word 认为 TI 在半导体电子方面的专业知识可以使空军的炸弹更加准确。德州仪器对设计炸弹一无所知,所以消息始于一枚标准型炸弹——750 磅重的 M-117,其中 638 枚已在清化桥附近投掷失败。他添加了一小组机翼,可以在炸弹从天而降时指挥炸弹的飞行。最后,他安装了一个简单的激光制导系统来控制机翼。一个小的硅晶片被分成四个象限并放在一个透镜后面。从目标反射的激光将通过透镜照射到硅上。如果炸弹偏离轨道,一个象限会比其他象限接收更多的激光能量,电路将移动机翼以重新调整炸弹的轨迹,使激光直接穿过透镜。

Word thought TI’s expertise in semiconductor electronics could make the Air Force’s bombs more accurate. Texas Instruments knew nothing about designing bombs, so Word started with a standard-issue bomb—the 750-pound M-117, 638 of which already had been dropped unsuccessfully around the Thanh Hoa Bridge. He added a small set of wings that could direct the bomb’s flight as it fell from the sky. Finally, he installed a simple laser-guidance system that would control the wings. A small silicon wafer was divided into four quadrants and placed behind a lens. The laser reflecting off the target would shine through the lens onto the silicon. If the bomb veered off course, one quadrant would receive more of the laser’s energy than the others, and circuitry would move the wings to reorient the bomb’s trajectory so that the laser was shining straight through the lens.

戴维斯上校给了德州仪器 9 个月和 99,000 美元来交付这种激光制导炸弹,由于其简单的设计,它很快通过了空军的测试。1972 年 5 月 13 日,美国飞机在清化大桥上投下了 24 枚炸弹,直到那一天,清化大桥仍然矗立在数百个陨石坑中,就像一座世纪中期轰炸战术不准确的纪念碑。这一次,美国的炸弹直接命中。数十座其他桥梁、铁路枢纽和其他战略要地被新型精确炸弹击中。一个简单的激光传感器和几个晶体管将一把 638 命中率为零的武器变成了精确破坏的工具。

Colonel Davis gave Texas Instruments nine months and $99,000 to deliver this laser-guided bomb, which, thanks to its simple design, quickly passed the Air Force’s tests. On May 13, 1972, U.S. aircraft dropped twenty-four of the bombs on the Thanh Hoa Bridge, which until that day had been still standing amid hundreds of craters, like a monument to the inaccuracy of mid-century bombing tactics. This time, American bombs scored direct hits. Dozens of other bridges, rail junctions, and other strategic points were hit with new precision bombs. A simple laser sensor and a couple of transistors had turned a weapon with a zero-for-638 hit ratio into a tool of precision destruction.

最终,越南农村的游击战争不是空袭能赢的。TI 宝石路激光制导炸弹的问世恰逢美国在战争中的失败。当像威廉威斯特摩兰将军这样的军事领导人预测“受到实时或接近实时监视的战斗区域”时,“自动射击控制,”许多人听到了最初将美国拖入越南的狂妄自大的回声。因此,除了少数军事理论家和电气工程师之外,几乎没有人意识到越南是一个成功的武器试验场,这些武器将微电子和爆炸物结合在一起,从而彻底改变战争并改变美国的军事实力。

In the end, the guerilla war in Vietnam’s countryside wasn’t a fight that aerial bombing could win. The arrival of TI’s Paveway laser-guided bombs coincided with America’s defeat in the war. When military leaders like General William Westmoreland predicted “combat areas that are under real- or near real-time surveillance” and “automated fire control,” many people heard echoes of the hubris that had dragged America into Vietnam in the first place. Outside a small number of military theorists and electrical engineers, therefore, hardly anyone realized Vietnam had been a successful testing ground for weapons that married microelectronics and explosives in ways that would revolutionize warfare and transform American military power.

第十二章 供应链治国之道

CHAPTER 12 Supply Chain Statecraft

尽管德州仪器 ( TI ) 执行官 Mark Shepherd 在二战期间曾在亚洲海军服役,但 Morris Chang 打趣说,他在该地区的专业知识并没有超出“酒吧和舞女。” 这Shepherd 是达拉斯警官的儿子,他在 6 岁时组装了他的第一个真空管。他在建立 TI 的半导体业务方面发挥了核心作用,包括监督 Jack Kilby 在发明第一个集成电路时所在的部门。Shepherd 有着宽阔的肩膀、直立的衣领、光滑的头发和紧绷的笑容,看起来就像他曾经是德克萨斯州的企业巨头。现在,他准备领导 TI 将部分生产外包到亚洲的战略。

Though Texas Instruments executive Mark Shepherd had served in the Navy in Asia during World War II, Morris Chang quipped that his expertise in the region didn’t extend beyond “bars and dancing girls.” The son of a Dallas police officer, Shepherd had assembled his first vacuum tube at age six. He’d played a central role in building TI’s semiconductor business, including supervising the division Jack Kilby worked in when the first integrated circuit was invented. With broad shoulders, a starched collar, slicked-back hair, and a taut smile, Shepherd looked like the Texas corporate titan that he was. Now he was poised to lead TI’s strategy of offshoring some of its production to Asia.

Chang 和 Shepherd 于 1968 年首次访问台湾,作为亚洲之行的一部分,为新的芯片组装厂选址。这次访问再糟糕不过了。当他的牛排配上酱油时,Shepherd 的反应很激烈,这与德克萨斯州通常的准备方式不同。他与台湾强大而精明的经济部长KT Li的第一次会面在部长宣布知识产权是“帝国主义者习惯于欺负欠发达国家。”

Chang and Shepherd first visited Taiwan in 1968 as part of an Asian tour to select a location for a new chip assembly facility. The visit couldn’t have gone worse. Shepherd reacted furiously when his steak was served with soy sauce, not the way it was usually prepared in Texas. His first meeting with Taiwan’s powerful and savvy economy minister, K. T. Li, ended acrimoniously when the minister declared that intellectual property was something “imperialists used to bully less-advanced countries.”

李认为牧羊人是美国帝国的代理人并没有错。但与试图将美国驱逐出境的北越人不同,李最终意识到台湾会受益于与美国更深入地融合。台湾和美国自 1955 年以来一直是条约盟友,但在越南战败后,美国的安全承诺看起来摇摇欲坠。从韩国到台湾,从马来西亚到新加坡,反共政府都在寻求保证,美国从越南撤军不会让他们孤立无援。他们也在寻找工作和投资,以解决驱使他们的一些人口走向共产主义的经济不满。李部长意识到德州仪器可以帮助台湾同时解决这两个问题。

Li wasn’t wrong to see Shepherd as an agent of America’s empire. But unlike the North Vietnamese, who were trying to oust the United States from their country, Li eventually realized that Taiwan would benefit from integrating itself more deeply with the United States. Taiwan and the U.S. had been treaty allies since 1955, but amid the defeat in Vietnam, America’s security promises were looking shaky. From South Korea to Taiwan, Malaysia to Singapore, anti-Communist governments were seeking assurance that America’s retreat from Vietnam wouldn’t leave them standing alone. They were also seeking jobs and investment that could address the economic dissatisfaction that drove some of their populations toward Communism. Minister Li realized that Texas Instruments could help Taiwan solve both problems at once.

在华盛顿,美国战略家担心美国支持的南越即将崩溃会在亚洲引发冲击波。外交政策战略家认为,该地区的华人社区已经成熟,可以像多米诺骨牌一样落入共产党的影响之下。例如,马来西亚的华裔少数民族是该国共产党的骨干。新加坡躁动不安的工人阶级以华裔居多。北京正在寻找盟友,并探寻美国的弱点。

In Washington, U.S. strategists feared the coming collapse of American-backed South Vietnam would send shock waves across Asia. Foreign policy strategists perceived ethnic Chinese communities all over the region as ripe for Communist penetration, ready to fall to Communist influence like a cascade of dominoes. Malaysia’s ethnic Chinese minority formed the backbone of that country’s Communist Party, for example. Singapore’s restive working class was majority ethnic Chinese. Beijing was searching for allies—and probing for U.S. weakness.

没有人比台湾政府更担心共产党在越南即将取得胜利,它仍然声称统治着整个中国。1960 年代对台湾经济来说是一个不错的十年,但对其外交政策却是灾难性的。岛上的独裁者蒋介石仍然梦想重新征服大陆,但军事平衡已经决定性地转向反对他。1964年,北京试验了第一颗原子弹。随后不久进行了热核武器试验。面对一个有核的中国,台湾比以往任何时候都更需要美国的安全保障。然而,随着越南战争的拖延,美国削减了对其亚洲朋友(包括台湾)的经济援助,这对于一个如此依赖美国支持的国家来说是一个不祥的迹象。

No one was more worried about the impending Communist victory in Vietnam than the government in Taiwan, which still claimed to rule all of China. The 1960s had been a good decade for Taiwan’s economy but disastrous for its foreign policy. The island’s dictator, Chiang Kai-shek, still dreamed of reconquering the mainland, but the military balance had shifted decisively against him. In 1964, Beijing tested its first atomic weapon. A thermonuclear weapon test shortly followed. Facing a nuclear China, Taiwan needed American security guarantees more than ever. Yet as the war in Vietnam dragged on, the U.S. cut economic aid for its friends in Asia, including in Taiwan, an ominous sign for a country so dependent on American support.

台湾像 KT Li 这样的官员,曾在剑桥学习核物理并经营过一家钢铁厂,然后在战后几十年指导台湾的经济发展,他们开始制定与美国经济一体化的战略。半导体是该计划的核心。李知道有很多愿意提供帮助的台裔美国半导体工程师。在达拉斯,Morris Chang 敦促 TI 的同事在台湾设立工厂。很多人后来形容大陆出生的张某“回到”台湾,但1968年是他第一次踏上台湾,自从逃离共产党接管中国以来一直住在美国。然而,张在斯坦福大学的两个博士同学来自台湾,他们说服他台湾有良好的商业环境,工资会保持在低水平。

Taiwanese officials like K. T. Li, who’d studied nuclear physics at Cambridge and ran a steel mill before steering Taiwan’s economic development through the postwar decades, began crystallizing a strategy to integrate economically with the United States. Semiconductors were at the center of this plan. Li knew there were plenty of Taiwanese-American semiconductor engineers willing to help. In Dallas, Morris Chang urged his colleagues at TI to set up a facility in Taiwan. Many people would later describe the mainland-born Chang as “returning” to Taiwan, but 1968 was the first time he stepped foot on the island, having lived in the U.S. since fleeing the Communist takeover of China. Two of Chang’s PhD classmates at Stanford were from Taiwan, however, and they convinced him the island had a favorable business climate and that wages would stay low.

在最初指责马克·谢泼德是帝国主义者之后,李部长很快改变了态度。他意识到与德州仪器的关系可以改变台湾的经济、建设工业和转让技术诀窍。与此同时,电子组装将促进其他投资,帮助台湾生产更多价值更高的商品。随着美国人对在亚洲的军事承诺越来越怀疑,台湾迫切需要多样化与美国的联系。对保卫台湾不感兴趣的美国人可能愿意保卫德州仪器。岛上的半导体工厂越多,与美国的经济联系越多,台湾就越安全。1968年7月,与台湾政府关系缓和后,TI 董事会批准在台湾建设新工厂。到 1969 年 8 月,这家工厂开始组装第一批设备。到 1980 年,它已经出货了第 10 亿台。

After initially accusing Mark Shepherd of being an imperialist, Minister Li quickly changed his tune. He realized a relationship with Texas Instruments could transform Taiwan’s economy, building industry and transferring technological know-how. Electronics assembly, meanwhile, would catalyze other investments, helping Taiwan produce more higher-value goods. As Americans grew skeptical of military commitments in Asia, Taiwan desperately needed to diversify its connections with the United States. Americans who weren’t interested in defending Taiwan might be willing to defend Texas Instruments. The more semiconductor plants on the island, and the more economic ties with the United States, the safer Taiwan would be. In July 1968, having smoothed over relations with the Taiwanese government, TI’s board of directors approved construction of the new facility in Taiwan. By August 1969, this plant was assembling its first devices. By 1980, it had shipped its billionth unit.

台湾并不是唯一一个认为半导体供应链可以促进经济增长和加强政治稳定的人。1973 年,新加坡领导人李光耀告诉美国总统理查德尼克松,他指望出口到在新加坡“解决失业”。在新加坡政府的支持下,TI 和美国国家半导体在新加坡建立了组装设施。许多其他芯片制造商紧随其后。到 1970 年代末,美国半导体公司在国际上雇佣了数万名工人,主要在韩国、台湾和东南亚。德克萨斯州和加利福尼亚州的芯片制造商、亚洲独裁者以及为许多亚洲半导体组装设施配备工作人员的通常是华裔工人之间出现了一个新的国际联盟。

Taiwan wasn’t alone in thinking that semiconductor supply chains could provide economic growth and bolster political stability. In 1973, Singapore’s leader Lee Kuan Yew told U.S. president Richard Nixon he was counting on exports to “sop up unemployment” in Singapore. With the Singapore government’s support, TI and National Semiconductors built assembly facilities in the city-state. Many other chipmakers followed. By the end of the 1970s, American semiconductor firms employed tens of thousands of workers internationally, mostly in Korea, Taiwan, and Southeast Asia. A new international alliance emerged between Texan and Californian chipmakers, Asian autocrats, and the often ethnic-Chinese workers who staffed many of Asia’s semiconductor assembly facilities.

半导体重塑了美国在该地区的朋友的经济和政治。曾经是政治激进主义滋生地的城市被勤奋的流水线工人改造了,他们乐于用失业或自给农业换取更高薪水的工厂工作。到 1980 年代初,电子工业占新加坡国民生产总值的 7% 和四分之一的制造业工作岗位。在电子产品生产中,60% 是半导体设备,其余大部分是没有半导体就无法工作的商品。在香港,电子制造业创造的就业机会比纺织业以外的任何行业都多。在马来西亚,槟城、吉隆坡和马六甲的半导体生产蓬勃发展,高薪电子组装工作。

Semiconductors recast the economies and politics of America’s friends in the region. Cities that had been breeding grounds for political radicalism were transformed by diligent assembly line workers, happy to trade unemployment or subsistence farming for better paying jobs in factories. By the early 1980s, the electronics industry accounted for 7 percent of Singapore’s GNP and a quarter of its manufacturing jobs. Of electronics production, 60 percent was semiconductor devices, and much of the rest was goods that couldn’t work without semiconductors. In Hong Kong, electronics manufacturing created more jobs than any sector except textiles. In Malaysia, semiconductor production boomed in Penang, Kuala Lumpur, and Melaka, with new manufacturing jobs providing work for many of the 15 percent of Malaysian workers who had left farms and moved to cities between 1970 and 1980. Such vast migrations are often politically destabilizing, but Malaysia kept its unemployment rate low with many relatively well-paid electronics assembly jobs.

从韩国到台湾,从新加坡到菲律宾,半导体组装设施的地图看起来很像美国在亚洲的军事基地地图。然而,即使在美国最终承认在越南战败并缩减其在该地区的军事存在之后,这些跨太平洋供应链仍然存在。到 1970 年代末期,美国在亚洲的盟友并没有倒在共产主义之手,反而与美国更加深入地融合在一起

From South Korea to Taiwan, Singapore to the Philippines a map of semiconductor assembly facilities looked much like a map of American military bases across Asia. Yet even after the U.S. finally admitted defeat in Vietnam and drew down its military presence in the region, these trans-Pacific supply chains endured. By the end of the 1970s, rather than dominoes falling to Communism, America’s allies in Asia were even more deeply integrated with the U.S.

1977 年,Mark Shepherd 回到台湾,在他们第一次见面近十年后再次与 KT Li 会面。台湾仍然面临中国入侵的风险,但谢泼德告诉李,“我们认为这种风险被台湾经济的实力和活力所抵消。TI 将留在台湾并继续在台湾发展,”他承诺。该公司今天在岛上仍有设施。与此同时,台湾已成为硅谷不可替代的合作伙伴。

In 1977, Mark Shepherd returned to Taiwan and met again with K. T. Li, nearly a decade after their first meeting. Taiwan still faced a risk of Chinese invasion, but Shepherd told Li, “We consider this risk to be more than offset by the strength and dynamism of Taiwan’s economy. TI will stay and continue to grow in Taiwan,” he promised. The company still has facilities on the island today. Taiwan, meanwhile, has made itself an irreplaceable partner to Silicon Valley.

第 13 章 英特尔的革命者

CHAPTER 13 Intel’s Revolutionaries

1968年似乎是一个革命性的时刻。从北京到柏林再到伯克利,激进分子和左翼分子蓄势待发,准备摧毁既定秩序。北越的春节攻势考验了美国军事力量的极限。然而,《帕洛阿尔托时报》在第 6 页报道了世界上最大的报纸,事后看来,这是一年中最具革命性的事件:“创始人离开仙童;组建自己的电子公司。”

The year 1968 seemed like a revolutionary moment. From Beijing to Berlin to Berkeley, radicals and leftists were poised to tear down the established order. North Vietnam’s Tet Offensive tested the limits of American military power. Yet it was the Palo Alto Times that scooped the world’s biggest newspapers by reporting on page 6 what, in hindsight, was the most revolutionary event of the year: “Founders Leave Fairchild; Form Own Electronics Firm.”

鲍勃·诺伊斯(Bob Noyce)和戈登·摩尔(Gordon Moore)的叛乱看起来不像加州东湾的抗议活动,伯克利的学生和黑豹党在那里策划暴力起义并梦想废除资本主义。在仙童,诺伊斯和摩尔对他们缺乏股票期权感到不满,并且厌倦了公司纽约总部的干预。他们的梦想不是推翻既定的秩序,而是重塑它。

The rebellion of Bob Noyce and Gordon Moore didn’t look like the protests in California’s East Bay, where Berkeley students and Black Panthers plotted violent uprisings and dreamt of abolishing capitalism. At Fairchild, Noyce and Moore were unhappy about their lack of stock options and sick of meddling from the company’s head office in New York. Their dream wasn’t to tear down the established order, but to remake it.

诺伊斯和摩尔像十年前离开肖克利的初创公司一样迅速地放弃了仙童,并创立了英特尔,它代表集成电子。在他们的愿景中,晶体管将成为有史以来最便宜的产品,但世界将消耗数万亿个晶体管。人类将被半导体赋予权力,同时从根本上依赖它们。即使世界正在与美国连线,美国的内部电路正在改变。工业时代即将结束。将晶体管蚀刻到硅中的专业知识现在将塑造世界经济。像帕洛阿尔托和山景城这样的加州小城镇正准备成为新的全球权力中心。

Noyce and Moore abandoned Fairchild as quickly as they’d left Shockley’s startup a decade earlier, and founded Intel, which stood for Integrated Electronics. In their vision, transistors would become the cheapest product ever produced, but the world would consume trillions and trillions of them. Humans would be empowered by semiconductors while becoming fundamentally dependent on them. Even as the world was being wired to the United States, America’s internal circuitry was changing. The industrial era was ending. Expertise in etching transistors into silicon would now shape the world’s economy. Small California towns like Palo Alto and Mountain View were poised to become new centers of global power.

成立两年后,英特尔推出了其第一款产品,一种称为动态随机存取存储器或 DRAM 的芯片。在 1970 年代之前,计算机通常不是使用硅芯片,而是使用一种称为磁芯的设备来“记住”数据,磁芯是一种由电线网格串在一起的微小金属环矩阵。当一个环被磁化时,它为计算机存储一个 1;未磁化的环是 0。将环串在一起的电线丛林可以关闭和打开每个环的磁性,并且可以“读取”给定环是 1 还是 0。记住 1 和 0 的需求是然而,爆炸,电线和环只能收缩到现在。如果组件变得更小,手工将它们编织在一起的装配工将发现它们无法生产。随着对计算机内存的需求激增,磁芯跟不上。

Two years after its founding, Intel launched its first product, a chip called a dynamic random access memory, or DRAM. Before the 1970s, computers generally “remembered” data using not silicon chips but a device called a magnetic core, a matrix of tiny metal rings strung together by a grid of wires. When a ring was magnetized, it stored a 1 for the computer; a non-magnetized ring was a 0. The jungle of wires that strung the rings together could turn each ring’s magnetism off and on and could “read” whether a given ring was a 1 or a 0. The demand for remembering 1s and 0s was exploding, however, and wires and rings could only shrink so far. If the components got any smaller, the assemblers who weaved them together by hand would find them impossible to produce. As demand for computer memory exploded, magnetic cores couldn’t keep up.

在 1960 年代,像 IBM 的 Robert Dennard 这样的工程师开始设想可以比小金属环更有效地“记忆”的集成电路。丹纳德有一头乌黑的长发从耳下垂下,直角直射,与地面平行,给人一种古怪天才的感觉。他建议将微型晶体管与电容器耦合,电容器是一种可以充电(1)或不充电(0)的微型存储设备。电容器会随着时间的推移而泄漏,因此 Dennard 设想通过晶体管反复为电容器充电。该芯片将被称为动态(由于重复充电)随机存取存储器或 DRAM。直到今天,这些芯片构成了计算机内存的核心。

In the 1960s, engineers like IBM’s Robert Dennard began envisioning integrated circuits that could “remember” more efficiently than little metal rings. Dennard had long, dark hair that flowed below his ears, then shot out at a right angle, parallel to the ground, giving him the look of an eccentric genius. He proposed coupling a tiny transistor with a capacitor, a miniature storage device that is either charged (1) or not (0). Capacitors leak over time, so Dennard envisioned repeatedly charging the capacitor via the transistor. The chip would be called a dynamic (due to the repeated charging) random access memory, or DRAM. These chips form the core of computer memory up to the present day.

DRAM 芯片像旧的磁芯存储器一样工作,借助电流存储 1 和 0。但是,DRAM 电路不是依靠电线和环,而是雕刻在硅中。它们不需要手工编织,因此故障较少,并且可以做得更小。诺伊斯和摩尔打赌他们的新公司英特尔可以将丹纳德的洞察力应用到芯片上比磁芯更致密。只看一眼摩尔定律的图表就知道,只要硅谷能够继续缩小晶体管,DRAM 芯片就会征服计算机内存业务。

A DRAM chip worked like the old magnetic core memories, storing 1s and 0s with the help of electric currents. But rather than relying on wires and rings, DRAM circuits were carved into silicon. They didn’t need to be weaved by hand, so they malfunctioned less often and could be made far smaller. Noyce and Moore bet that their new company, Intel, could take Dennard’s insight and put it on a chip far denser than a magnetic core could ever be. It only took one glance at a graph of Moore’s Law to know that so long as Silicon Valley could keep shrinking transistors, DRAM chips would conquer the business of computer memory.

英特尔计划主导 DRAM 芯片业务。存储芯片不需要专门化,因此具有相同设计的芯片可以用于许多不同类型的设备。这使得大批量生产它们成为可能。相比之下,另一种主要类型的芯片——那些负责“计算”而不是“记忆”的芯片——是专门为每种设备设计的,因为每个计算问题都是不同的。例如,计算器的工作方式与导弹的制导计算机不同,因此直到 1970 年代,他们使用不同类型的逻辑芯片。这种专业化推高了成本,因此英特尔决定专注于内存芯片,大规模生产将产生规模经济。

Intel planned to dominate the business of DRAM chips. Memory chips don’t need to be specialized, so chips with the same design can be used in many different types of devices. This makes it possible to produce them in large volumes. By contrast, the other main type of chips—those tasked with “computing” rather than “remembering”—were specially designed for each device, because every computing problem was different. A calculator worked differently than a missile’s guidance computer, for example, so until the 1970s, they used different types of logic chips. This specialization drove up cost, so Intel decided to focus on memory chips, where mass production would produce economies of scale.

然而,鲍勃·诺伊斯永远无法抗拒工程难题。尽管他刚刚为他的新公司将生产存储芯片的承诺筹集了数百万美元,但他很快就被说服增加一条产品线。1969 年,一家名为 Busicom 的日本计算器公司找到诺伊斯,要求为其最新的计算器设计一套复杂的电路。手持计算器是 1970 年代的 iPhone,该产品使用最先进的计算技术来降低价格,并将一块强大的塑料放入每个人的口袋。许多日本公司制造了计算器,但他们经常依靠硅谷来设计和制造他们的芯片。

Bob Noyce could never resist an engineering puzzle, however. Even though he’d just raised several million dollars on the promise that his new company would build memory chips, he was quickly convinced to add a product line. In 1969, a Japanese calculator firm called Busicom approached Noyce with a request to design a complicated set of circuits for its newest calculator. Handheld calculators were the iPhones of the 1970s, a product that used the most advanced computing technologies to drive down price and put a powerful piece of plastic in everyone’s pocket. Many Japanese firms built calculators, but they often relied on Silicon Valley to design and manufacture their chips.

Noyce 请 Ted Hoff 处理 Busicom 的请求,他是一位说话温和的工程师,他在完成神经网络的学术生涯后来到英特尔。与大多数英特尔员工不同的是,他们是物理学家或化学家,专注于电子在芯片上的快速移动,霍夫在计算机架构方面的背景让他从半导体驱动的系统的角度来看待半导体。Busicom 告诉 Hoff,他们需要 12 种不同的芯片和 24000 个晶体管,全部按定制设计排列。他认为对于像英特尔这样的小型初创公司来说,这听起来非常复杂。

Noyce asked Ted Hoff, a soft-spoken engineer who’d arrived at Intel after an academic career studying neural networks, to handle Busicom’s request. Unlike most Intel employees, who were physicists or chemists focused on the electrons zipping across chips, Hoff’s background in computer architectures let him see semiconductors from the perspective of the systems they powered. Busicom told Hoff they’d need twelve different chips with twenty-four thousand transistors, all arranged in a bespoke design. He thought this sounded impossibly complicated for a small startup like Intel.

在考虑 Busicom 的计算器时,霍夫意识到计算机面临定制逻辑电路和定制软件之间的权衡。因为芯片制造是一项定制业务,为每个设备提供专门的电路,所以客户不会认真考虑软件。然而,英特尔在内存芯片方面的进步——以及随着时间的推移它们将变得更加强大的前景——意味着计算机很快就会拥有处理复杂软件所需的内存容量。霍夫打赌,设计一个标准化的逻辑芯片,再加上一个用不同类型软件编程的强大存储芯片,可以计算许多不同的东西,很快就会更便宜。毕竟,霍夫知道,没有人制造出比英特尔更强大的内存芯片。

As he considered Busicom’s calculator, Hoff realized computers face a tradeoff between customized logic circuits and customized software. Because chipmaking was a custom business, delivering specialized circuits for each device, customers didn’t think hard about software. However, Intel’s progress with memory chips—and the prospect they would become exponentially more powerful over time—meant computers would soon have the memory capacity needed to handle complex software. Hoff bet it would soon be cheaper to design a standardized logic chip that, coupled with a powerful memory chip programmed with different types of software, could compute many different things. After all, Hoff knew, no one was building memory chips more powerful than Intel’s.

英特尔并不是第一家考虑生产通用逻辑芯片的公司。一家国防承包商为 F-14 战斗机上的计算机生产了一种类似于英特尔的芯片。然而,该芯片的存在直到 1990 年代才被保密。然而,英特尔推出了一款名为 4004 的芯片,并将其描述为世界上第一款微处理器——正如该公司的广告宣传所言,“芯片上的微型可编程计算机”。它可以用于许多不同类型的设备和掀起了一场计算革命。

Intel wasn’t the first company to think about producing a generalized logic chip. A defense contractor had produced a chip much like Intel’s for the computer on the F-14 fighter jet. However, that chip’s existence was kept secret until the 1990s. Intel, however, launched a chip called the 4004 and described it as the world’s first microprocessor—“a micro-programmable computer on a chip,” as the company’s advertising campaign put it. It could be used in many different types of devices and set off a revolution in computing.

1972 年,在他父母的五十周年结婚纪念日派对上,鲍勃·诺伊斯打断了庆祝活动,举起一块硅片,向家人宣布:“这将改变世界。” 现在可以大规模生产通用逻辑。计算已经为自己的工业革命做好了准备,英特尔拥有世界上最先进的装配线。

At his parents’ fiftieth wedding anniversary party in 1972, Bob Noyce interrupted the festivities, held up a silicon wafer, and declared to his family: “This is going to change the world.” Now general logic could be mass-produced. Computing was ready for its own industrial revolution and Intel had the world’s most advanced assembly lines.

最了解大规模生产的计算能力将如何彻底改变社会的人是加州理工学院的一位名叫 Carver Mead 的教授。眼神犀利,留着山羊胡,米德看起来更像是伯克利的哲学家,而不是电气工程师。他与戈登摩尔建立了友谊,就在仙童公司成立后,摩尔华尔兹进入米德的加州理工学院办公室,拿出一只装满雷神 2N706 晶体管的袜子,送给米德在他的电气工程课上使用。摩尔很快聘请米德担任顾问,并多年来,这位加州理工学院的梦想家每周三都在英特尔位于硅谷的工厂度过。尽管 Gordon Moore 在他 1965 年的著名文章中首次描绘了晶体管密度的指数增长,但 Mead 创造了术语“摩尔定律”来描述它。

The person who best understood how mass-produced computing power would revolutionize society was a Caltech professor named Carver Mead. With piercing eyes and a goatee, Mead looked more like a Berkeley philosopher than an electrical engineer. He had struck up a friendship with Gordon Moore just after the founding of Fairchild, after Moore waltzed into Mead’s Caltech office, pulled out a sock filled with Raytheon 2N706 transistors, and gave them to Mead for use in his electrical engineering classes. Moore soon hired Mead as a consultant, and for many years, the Caltech visionary spent each Wednesday at Intel’s facilities in Silicon Valley. Though Gordon Moore had first graphed the exponential increase in transistor density in his famous 1965 article, Mead coined the term “Moore’s Law” to describe it.

“未来十年,”米德在 1972 年预测,“我们社会的方方面面都将在某种程度上实现自动化。” 他设想“在我们的电话、洗衣机或汽车内部深处有一台微型计算机”,因为这些硅芯片变得无处不在且价格低廉。“在过去的 200 年里,我们将制造商品和运送人员的能力提高了 100 倍,”米德计算道。“但在过去 20 年中,我们处理和检索信息的速度增加了 1,000,000 到 10,000,000。” 数据处理的革命性爆炸即将到来。“我们有计算机能力从我们的耳朵里冒出来。”

“In the next ten years,” Mead predicted in 1972, “every facet of our society will be automated to some degree.” He envisioned “a tiny computer deep down inside of our telephone, or our washing machine, or our car” as these silicon chips became pervasive and inexpensive. “In the past 200 years we have improved our ability to manufacture goods and move people by a factor of 100,” Mead calculated. “But in the last 20 years there has been an increase of 1,000,000 to 10,000,000 in the rate at which we process and retrieve information.” A revolutionary explosion of data processing was coming. “We have computer power coming out of our ears.”

米德预言了一场具有深远社会和政治后果的革命。在这个新世界中的影响力将累积到能够产生计算能力并用软件操纵它的人身上。硅谷的半导体工程师拥有专业知识、网络和股票期权,可以让他们制定未来的规则——其他人都必须遵守的规则。工业社会正在让位于数字世界,1 和 0 在遍布整个社会的数百万块硅片上存储和处理。科技大亨的时代即将来临。“社会的命运将悬而未决,”Carver Mead 宣称。“催化剂是微电子技术及其将越来越多的组件放入越来越小的空间的能力。” 行业外人只是模糊地感知到世界正在如何变化,但英特尔的领导者知道,如果他们成功地大幅扩展了计算能力的可用性,就会发生根本性的变化。“我们真的是当今世界的革命者,”戈登·摩尔在 1973 年宣称,“而不是几年前毁坏学校的留着长发和胡须的孩子。”

Mead was prophesying a revolution with profound social and political consequences. Influence in this new world would accrue to people who could produce computing power and manipulate it with software. The semiconductor engineers of Silicon Valley had the specialized knowledge, networks, and stock options that let them write the rules of the future—rules everyone else would have to follow. Industrial society was giving way to a digital world, with 1s and 0s stored and processed on many millions of slabs of silicon spread throughout society. The era of the tech tycoons was dawning. “Society’s fate will hang in the balance,” Carver Mead declared. “The catalyst is the microelectronics technology and its ability to put more and more components into less and less space.” Industry outsiders only dimly perceived how the world was changing, but Intel’s leaders knew that if they succeeded in drastically expanding the availability of computing power, radical changes would follow. “We are really the revolutionaries in the world today,” Gordon Moore declared in 1973, “not the kids with the long hair and beards who were wrecking the schools a few years ago.”

第 14 章 五角大楼的抵消战略

CHAPTER 14 The Pentagon’s Offset Strategy

没有人从诺伊斯和摩尔的革命中得到的好处多于角落里的人旧秩序之一——五角大楼。1977 年抵达华盛顿后,威廉·佩里感觉“就像糖果店里的孩子”。他说,对于佩里这样的硅谷企业家来说,担任负责研究和工程的国防部副部长是“世界上最好的工作”。没有人比五角大楼拥有更多购买技术的预算。华盛顿几乎没有人如此清楚地了解微处理器和强大的存储芯片如何改变国防部所依赖的所有武器和系统。

No one benefitted more from Noyce and Moore’s revolution than a cornerstone of the old order—the Pentagon. Upon arriving in Washington in 1977, William Perry felt “like a kid in a candy store.” For a Silicon Valley entrepreneur like Perry, serving as undersecretary of defense for research and engineering was, he said, the “best job in the world.” No one had a larger budget to buy technology than the Pentagon. And hardly anyone in Washington had so clear a view of how microprocessors and powerful memory chips could transform all the weapons and systems the Defense Department relied on.

与鲍勃·诺伊斯(Bob Noyce)或戈登·摩尔(Gordon Moore)通过无视政府并为大众市场计算器和大型计算机销售芯片而发家致富不同,佩里对五角大楼非常了解。作为宾夕法尼亚州面包师的儿子,他的职业生涯始于硅谷科学家,为 Sylvania 电子防御实验室工作,该实验室隶属于同一家电子公司,该公司在他毕业于麻省理工学院后聘用了 Morris Chang。Perry 在加利福尼亚州的 Sylvania 工作,其任务是设计高度机密的电子设备,以监控苏联导弹的发射。1963 年秋天,他是被紧急召集到华盛顿检查 U-2 间谍飞机拍摄的新照片的十位专家之一在古巴展示苏联导弹。佩里很小的时候就被视为该国最顶尖的军事专家之一。

Unlike Bob Noyce or Gordon Moore, who were making a fortune by ignoring the government and selling chips for mass market calculators and mainframe computers, Perry knew the Pentagon intimately. The son of a Pennsylvania baker, he began his career as a Silicon Valley scientist working for Sylvania Electronic Defense Laboratories, a unit of the same electronics company that had hired Morris Chang after he graduated MIT. Working for Sylvania in California, Perry was tasked with designing highly classified electronics that monitored Soviet missile launches. In fall 1963, he’d been one of ten experts urgently called to Washington to examine new photographs taken by U-2 spy planes showing Soviet missiles in Cuba. At a young age, Perry was already seen as one of the country’s top experts on military affairs.

佩里在 Sylvania 的工作使他一跃成为美国的国防机构。但他仍然住在山景城。对于一个被初创公司包围的工程师来说,老式的 Sylvania 开始显得官僚和乏味。它的技术很快就过时了。在硅谷的芯片制造商大量生产集成电路之后很久,它的消费和军用产品都依赖真空管。佩里非常熟悉他身边的固态电子技术的进步。他和鲍勃·诺伊斯在同一个帕洛阿尔托牧歌合唱团演唱。因此,1963 年,佩里感受到正在进行的革命,独自出发,成立了自己的公司,为军队设计监控设备。为了获得他需要的处理能力,佩里从他的歌唱伙伴,英特尔的 CEO 那里购买了芯片。

Perry’s job at Sylvania had catapulted him into America’s defense establishment. But he still lived in Mountain View. For an engineer surrounded by startups, old-school Sylvania began to seem bureaucratic and stodgy. Its technology was quickly becoming outdated. Its consumer and military products alike relied on vacuum tubes long after Silicon Valley’s chipmakers were churning out integrated circuits. Perry was intimately familiar with the advances in solid-state electronics all around him. He sang in the same Palo Alto madrigals choir as Bob Noyce. So, sensing the revolution that was underway, in 1963, Perry had set off on his own, founding his own firm to design surveillance devices for the military. To get the processing power he needed, Perry bought chips from his singing partner, Intel’s CEO.

在阳光明媚的硅谷,感觉就像“一切都是新的,一切皆有可能”,佩里后来回忆道。在他 1977 年抵达五角大楼时,从五角大楼看,世界看起来要黑暗得多。美国刚刚输掉了越南战争。更糟糕的是,苏联几乎完全侵蚀了美国的军事优势,像安德鲁·马歇尔这样的五角大楼分析师警告说。马歇尔出生在底特律,身材矮小,秃顶,喙鼻,透过眼镜不可思议地凝视着这个世界。他曾在一家二战期间的机床厂,后来成为上半个世纪最有影响力的政府官员之一。马歇尔于 1973 年受聘建立五角大楼网络评估办公室,负责预测战争的未来。

In sunny Silicon Valley it felt like “everything was new and anything was possible,” Perry would later remember. Viewed from the Pentagon upon his arrival in 1977, the world looked far darker. The U.S. had just lost the Vietnam War. Worse, the Soviet Union had almost completely eroded America’s military advantage, warned Pentagon analysts like Andrew Marshall. Born in Detroit, Marshall was a small man, with a bald head and a beaky nose, who stared inscrutably at the world from behind his glasses. He’d worked in a machine tools factory during World War II, before becoming one of the most influential government officials of the last half century. Marshall had been hired in 1973 to establish the Pentagon’s Office of Net Assessment and was tasked with forecasting the future of war.

马歇尔的严峻结论是,在东南亚进行了十年毫无意义的战斗之后,美国已经失去了军事优势。他执着于重新获得它。尽管华盛顿对人造卫星和古巴导弹危机感到震惊,但直到 1970 年代初,苏联才建立了足够大的洲际弹道导弹储备,以保证他们的原子武器能够在美国的核打击中幸存下来以进行报复带有毁灭性的原子自己的弹幕。更令人担忧的是,苏联军队拥有更多的坦克和飞机,它们已经部署在欧洲的潜在战场上。美国——在国内面临削减军费开支的压力——根本跟不上。

Marshall’s grim conclusion was that after a decade of pointless fighting in Southeast Asia, the U.S. had lost its military advantage. He was fixated on regaining it. Though Washington had been shocked by Sputnik and the Cuban Missile Crisis, it wasn’t until the early 1970s that the Soviets had built a big enough stockpile of intercontinental ballistic missiles to guarantee that enough of their atomic weapons could survive a U.S. nuclear strike to retaliate with a devastating atomic barrage of their own. More worrisome, the Soviet army had far more tanks and planes, which were already deployed on potential battlegrounds in Europe. The U.S.—facing pressure at home to cut military spending—simply couldn’t keep up.

像马歇尔这样的战略家知道,苏联数量优势的唯一答案是生产质量更好的武器。但是怎么做?早在 1972 年,马歇尔就写道,美国需要利用其计算机中的“大量和持久的铅”。他写道:“一个好的策略是发展这种领先优势,并以利用它的方式转变战争概念。” 他设想了导弹的“快速信息收集”、“复杂的指挥和控制”和“终端制导”,想象出能够以几乎完美的精度打击目标的弹药。如果战争的未来变成了一场准确性的较量,马歇尔打赌,苏联人会落后。

Strategists like Marshall knew the only answer to the Soviet quantitative advantage was to produce better quality weapons. But how? As early as 1972, Marshall wrote that the U.S. needed to take advantage of its “substantial and durable lead” in computers. “A good strategy would be to develop that lead and to shift concepts of warfare in ways that capitalize on it,” he wrote. He envisioned “rapid information gathering,” “sophisticated command and control,” and “terminal guidance” for missiles, imagining munitions that could strike targets with almost perfect accuracy. If the future of war became a contest for accuracy, Marshall wagered, the Soviets would fall behind.

佩里意识到,由于计算能力的小型化,马歇尔对战争未来的设想很快就会成为可能。他非常熟悉硅谷的半导体创新,曾在公司自己的设备中使用过英特尔的芯片。越南战争中使用的许多武器系统仍然依赖真空管,但最新的手持计算器中的芯片提供的计算能力比旧的麻雀 III 导弹要强大得多。佩里打赌,把这些筹码放在导弹上,美国军队就会领先于苏联。

Perry realized that Marshall’s vision of the future of war would soon be possible due to the miniaturization of computing power. He was intimately familiar with Silicon Valley’s semiconductor innovation, having used Intel’s chips in his company’s own devices. Many of the weapons systems used in the Vietnam War still relied on vacuum tubes, but chips in the newest handheld calculators offered vastly more computing power than an old Sparrow III missile. Put those chips in missiles, Perry wagered, and America’s military would jump ahead of the Soviets.

他推断,制导导弹不仅会“抵消”苏联的数量优势。他们会迫使苏联人采取极其昂贵的反导弹努力作为回应。佩里计算莫斯科需要五到十年30 到 500 亿美元用于防御五角大楼计划部署的 3000 枚美国巡航导弹——即便如此,如果导弹全部朝苏联发射,苏联也只能摧毁一半来袭导弹。

Guided missiles would not only “offset” the USSR’s quantitative advantage, he reasoned. They’d force the Soviets to undertake a ruinously expensive anti-missile effort in response. Perry calculated Moscow would need five to ten years and $30 to $50 billion to defend against the three thousand American cruise missiles that the Pentagon planned to field—and even then, the Soviets could only destroy half the incoming missiles if they were all fired at the USSR.

这正是 Andrew Marshall 一直在寻找的技术类型。与吉米·卡特的国防部长哈罗德·布朗、佩里和马歇尔合作,推动五角大楼大力投资于新技术:新一代制导导弹集成电路,而不是真空管;可以将位置坐标传送到地球上任何一点的卫星星座;还有——最重要的是——启动下一代芯片的新计划,以确保美国保持其技术优势。

This was exactly the type of technology that Andrew Marshall had been looking for. Working with Jimmy Carter’s secretary of defense, Harold Brown, Perry and Marshall pushed the Pentagon to invest heavily in new technologies: a new generation of guided missiles that used integrated circuits, not vacuum tubes; a constellation of satellites that could beam location coordinates to any point on earth; and—most important—a new program to jump-start the next generation of chips, to ensure that the U.S. kept its technological edge.

在佩里的带领下,五角大楼向利用美国在微电子领域的优势的新型武器系统投入了大量资金。诸如宝石路之类的精确武器计划得到了推广,从巡航导弹到炮弹的所有类型的制导弹药也得到了推广。传感器和通信也随着小型化计算能力的应用开始飞跃。例如,探测敌方潜艇主要是开发精确传感器并运行他们通过越来越复杂的算法收集的信息的问题。有了足够的处理能力,军方的声学专家打赌,应该可以从数英里外区分鲸鱼和潜艇。

Led by Perry, the Pentagon poured money into new weapons systems that capitalized on America’s advantage in microelectronics. Precision weapons programs like the Paveway were promoted, as were guided munitions of all types, from cruise missiles to artillery shells. Sensors and communications also began to leap forward with the application of miniaturized computing power. Detecting enemy submarines, for example, was largely a problem of developing accurate sensors and running the information they gathered through ever-more-complicated algorithms. With enough processing power, the military’s acoustic experts wagered, it should be possible to distinguish a whale from a submarine from many miles away.

制导武器变得更加复杂。像战斧导弹这样的新系统依赖于比宝石路更复杂的制导系统,使用雷达高度计扫描地面并将其与预加载到导弹计算机中的地形图相匹配。这样,如果导弹偏离航向,它就可以重新定向。这种类型的制导早在几十年前就已经被理论化了,但只有在强大的芯片小到可以装入巡航导弹的情况下才有可能实施。

Guided weaponry became more complex. New systems like the Tomahawk missile relied on far more sophisticated guidance systems than the Paveway, using a radar altimeter to scan the ground and match it with terrain maps preloaded into the missile’s computer. This way, the missile could redirect itself if it veered off course. This type of guidance had been theorized decades earlier but was only possible to implement now that powerful chips were small enough to fit in a cruise missile.

单个制导弹药是一项强大的创新,但如果它们能够共享信息,它们的影响力会更大。佩里委托了一项由五角大楼国防高级研究计划局 (DARPA) 运行的特殊计划,以了解如果将所有这些新传感器、制导武器和通信设备集成在一起会发生什么。称为“Assault Breaker”,它设想了一种可以识别敌方目标并向地面处理中心提供位置信息的空中雷达,该处理中心将把雷达细节与来自其他传感器的信息融合在一起。地基导弹将与空中雷达通信,引导它们朝着目标前进。在最后下降时,导弹将释放子弹药,这些子弹药将单独瞄准目标。

Individual guided munitions were a powerful innovation, but they’d be even more impactful if they could share information. Perry commissioned a special program, run via the Pentagon’s Defense Advanced Research Projects Agency (DARPA), to see what would happen if all these new sensors, guided weapons, and communications devices were integrated. Called “Assault Breaker,” it envisioned an aerial radar that could identify enemy targets and provide location information to a ground-based processing center, which would fuse the radar details with information from other sensors. Ground-based missiles would communicate with the aerial radar guiding them toward the target. On final descent, the missiles would release submunitions that would individually home in on their targets.

制导武器正在让位于自动化战争的愿景,计算能力以前所未有的方式分配给各个系统。正如佩里在 1981 年对一位采访者说的那样,美国正走在“将芯片密度提高 10 倍到 100 倍”的轨道上,并承诺在计算能力方面取得可比的增长,这才有可能。“我们将能够将仅在十年前填满整个房间的计算机放在芯片上”和领域“各级‘智能’武器。”

Guided weapons were giving way to a vision of automated war, with computing power distributed to individual systems in a way never before imaginable. This was only possible because the U.S. was on track “to increase the density of chips ten to a hundredfold,” as Perry told an interviewer in 1981, promising comparable increases in computing power. “We will be able to put computers, which only ten years ago would have filled up this entire room, on a chip” and field “ ‘smart’ weapons at all levels.”

佩里的愿景与硅谷所创造的任何东西一样激进。五角大楼真的可以实施一项高科技计划吗?到佩里 1981 年卸任时,卡特总统任期结束,记者和国会议员开始攻击他的精确打击赌博。“巡航导弹:神奇武器还是哑弹?” 一位专栏作家在 1983 年问道。另一位专栏作家将佩里的先进技术等同于“花里胡哨”,指出表面上“智能”的武器(如真空管动力麻雀导弹)的频繁故障和令人沮丧的杀伤率。

Perry’s vision was as radical as anything Silicon Valley had cooked up. Could the Pentagon really implement a high-tech program? By the time Perry left office in 1981, as the Carter presidency ended, journalists and members of Congress were attacking his gamble on precision strike. “Cruise Missiles: Wonder Weapon or Dud?” asked one columnist in 1983. Another equated Perry’s advanced technologies with “bells and whistles,” pointing out the frequent malfunctions and dismal kill ratio of ostensibly “smart” weapons like the vacuum tube−powered Sparrow missile.

佩里的愿景所需的计算能力进步对许多评论家来说似乎是科幻小说,他们认为导弹技术会缓慢改进,因为坦克和飞机也变化缓慢。摩尔定律规定的指数增长很少见,也很难理解。然而,佩里并不是唯一一个预测“十到一百倍”改进的人。英特尔向其客户承诺同样的事情。佩里抱怨他的国会批评者是“Luddites”,他们根本不了解芯片的变化速度有多快。

The advances in computing power that Perry’s vision required seemed like science fiction to many critics, who assumed guided missile technology would improve slowly because tanks and planes changed slowly, too. Exponential increases, which Moore’s Law dictated, are rarely seen and hard to comprehend. However, Perry wasn’t alone in predicting a “ten to a hundredfold” improvement. Intel was promising the very same thing to its customers. Perry grumbled that his congressional critics were “Luddites,” who simply didn’t understand how rapidly chips were changing.

即使在佩里离任后,国防部仍继续向先进芯片及其驱动的军事系统投入大量资金。Andrew Marshall 继续他在五角大楼的工作,他已经梦想着这些下一代芯片将使新系统成为可能。半导体工程师能否实现佩里承诺的进展?摩尔定律预测他们可以——但这只是一个预测,而不是保证。此外,与最初发明集成电路时不同,芯片行业已不再专注于军事生产。英特尔等公司的目标是企业计算机和消费品,而不是导弹。只有消费市场有足够的资金为摩尔定律所要求的庞大研发项目提供资金。

Even after Perry left office, the Defense Department continued to pour money into advanced chips and the military systems they powered. Andrew Marshall continued his work at the Pentagon, already dreaming of the new systems these next-generation chips would make possible. Could semiconductor engineers deliver the progress Perry promised? Moore’s Law predicted that they could—but this was only a prediction, not a guarantee. Moreover, unlike when integrated circuits were first invented, the chip industry had become less focused on military production. Firms like Intel targeted corporate computers and consumer goods, not missiles. Only consumer markets had the volume to fund the vast R&D programs that Moore’s Law required.

在 1960 年代初期,有可能声称五角大楼创造了硅谷。从那以后的十年里,形势发生了翻天覆地的变化。美国军队在越南输掉了战争,但芯片行业赢得了随后的和平,通过迅速扩大的投资联系和供应链,将亚洲其他地区从新加坡到台湾再到日本,与美国更加紧密地联系在一起。整个世界与美国的创新基础设施联系更加紧密,甚至像苏联这样的对手也花时间复制美国的芯片和芯片制造工具。与此同时,芯片行业催生了一系列新的武器系统,这些系统正在重塑美军未来战争的作战方式。美国的力量正在被重塑。现在整个国家都依赖硅谷的成功。

In the early 1960s, it had been possible to claim the Pentagon had created Silicon Valley. In the decade since, the tables had turned. The U.S. military lost the war in Vietnam, but the chip industry won the peace that followed, binding the rest of Asia, from Singapore to Taiwan to Japan, more closely to the U.S. via rapidly expanding investment links and supply chains. The entire world was more tightly connected to America’s innovation infrastructure, and even adversaries like the USSR spent their time copying U.S. chips and chipmaking tools. Meanwhile, the chip industry had catalyzed an array of new weapons systems that were remaking how the U.S. military would fight future wars. American power was being recast. Now the entire nation depended on Silicon Valley’s success.

第三部分 失去领导力?

PART III LEADERSHIP LOST?

第 15 章 “那场比赛很艰难”

CHAPTER 15 “That Competition Is Tough”

自从你写了那篇论文我的生活简直就是地狱!” 一位芯片销售员向惠普高管理查德安德森抱怨,他的任务是确定哪些芯片符合惠普的严格标准。1980 年代是整个美国半导体行业地狱般的十年。硅谷认为它在世界科技行业处于领先地位,但在经历了 20 年的快速增长之后,它现在面临着生存危机:来自日本的激烈竞争。1980 年 3 月 25 日,当安德森在华盛顿特区历史悠久的五月花酒店举行的行业会议上登台时,观众们仔细聆听,因为每个人都想向他推销他们的筹码。他工作的公司惠普在 1930 年代发明了硅谷创业公司的概念,当时斯坦福大学毕业生戴夫帕卡德和比尔休利特开始在帕洛阿尔托的车库里修补电子设备。

“Ever since you’ve written that paper, my life has been hell!” one chip salesman grumbled to Richard Anderson, a Hewlett-Packard executive tasked with deciding which chips met HP’s stringent standards. The 1980s were a hellish decade for the entire U.S. semiconductor sector. Silicon Valley thought it sat atop the world’s tech industry, but after two decades of rapid growth it now faced an existential crisis: cutthroat competition from Japan. When Anderson took the stage at an industry conference at Washington, D.C.’s historic Mayflower Hotel on March 25, 1980, the audience listened carefully, because everyone was trying to sell him their chips. Hewlett-Packard, the company he worked for, had invented the concept of a Silicon Valley startup in the 1930s, when Stanford grads Dave Packard and Bill Hewlett began tinkering with electronic equipment in a Palo Alto garage. Now it was one of America’s biggest tech companies—and one of the largest buyers of semiconductors.

安德森对芯片的判断可能会决定任何一家半导体公司的命运,但硅谷的销售人员从未被允许与他喝酒吃饭。“有时我让他们带我出去吃午饭,”他不好意思地承认。但整个山谷都知道,他是几乎每个人最重要客户的看门人。他的工作让他对半导体行业有一个全面的了解,包括每家公司的表现。

Anderson’s judgment about a chip could shape the fate of any semiconductor company, but Silicon Valley’s salesmen were never allowed to wine and dine him. “Sometimes I let them take me out to lunch,” he admitted sheepishly. But the entire valley knew that he was the gatekeeper to almost everyone’s most important customer. His job gave him a panoramic view of the semiconductor industry, including how each company was performing.

除了英特尔和德州仪器等美国公司外,东芝和 NEC 等日本公司现在也在制造 DRAM 内存芯片——尽管硅谷的大多数人并没有认真对待这些玩家。美国芯片制造商由发明高科技的人经营。他们开玩笑说日本是“咔嚓咔嚓”——日本工程师在芯片会议上为更好地复制想法而带来的相机发出的声音。美国主要芯片制造商卷入与日本竞争对手的知识产权诉讼这一事实被解释为硅谷仍然遥遥领先的证据。

In addition to American companies like Intel and TI, Japanese firms like Toshiba and NEC were now building DRAM memory chips—though most people in Silicon Valley didn’t take these players seriously. U.S. chipmakers were run by the people who’d invented high-tech. They joked that Japan was the country of “click, click”—the sound made by cameras that Japanese engineers brought to chip conferences to better copy the ideas. The fact that major American chipmakers were embroiled in intellectual property lawsuits with Japanese rivals was interpreted as evidence that Silicon Valley was still well ahead.

然而,在惠普,安德森并没有简单地认真对待东芝和 NEC——他测试了他们的芯片,发现它们的质量远远优于美国竞争对手。他报告说,这三家日本公司均未报告在其最初一千小时使用期间的故障率超过 0.02%。这三家美国公司的最低故障率为 0.09%——这意味着美国制造的芯片发生故障的数量是其四倍半。最差的美国公司生产的芯片故障率为 0.26%——超过10 比日本的成绩差几倍。美国 DRAM 芯片的工作原理相同,成本相同,但故障频率更高。那么为什么有人要买它们呢?

At HP, however, Anderson didn’t simply take Toshiba and NEC seriously—he tested their chips and found that they were of far better quality than American competitors. None of the three Japanese firms reported failure rates above 0.02 percent during their first one thousand hours of use, he reported. The lowest failure rate of the three American firms was 0.09 percent—which meant four-and-a-half times as many U.S.-made chips were malfunctioning. The worst U.S. firm produced chips with 0.26 percent failure rates—over ten times as bad as the Japanese results. American DRAM chips worked the same, cost the same, but malfunctioned far more often. So why should anyone buy them?

芯片并不是唯一面临来自高品质、超高效的日本竞争对手压力的美国行业。在战后不久,“日本制造”一直是“廉价”的代名词。但像索尼的 Akio Morita 这样的企业家已经摆脱了低价的名声,取而代之的是与任何美国竞争对手一样高品质的产品。森田的晶体管收音机是美国经济领先地位的第一个显着挑战者,他们的成功使森田和他的日本同行更加大胆地把目光投向了更高的地方。从汽车到钢铁的美国工业正面临日本的激烈竞争。

Chips weren’t the only U.S. industry facing pressure from high-quality, ultra-efficient Japanese competitors. In the immediate postwar years, “Made in Japan” had been a synonym for “cheap.” But entrepreneurs like Sony’s Akio Morita had cast off this reputation for low price, replacing it with products that were as high quality as those of any American competitor. Morita’s transistor radios were the first prominent challenger to American economic preeminence, and their success emboldened Morita and his Japanese peers to set their sights even higher. American industries from cars to steel were facing intense Japanese competition.

到 1980 年代,消费电子产品已成为日本的专长,索尼在推出新的消费品方面处于领先地位,从美国竞争对手手中抢占了市场份额。起初,日本公司通过复制美国竞争对手的产品取得成功,以更高的质量制造它们和更低的价格。一些日本人夸大了他们擅长实施的想法,而美国则更擅长创新。“我们没有诺伊斯博士或肖克利博士,”一位日本记者写道,尽管该国已经开始积累其诺贝尔奖获得者的份额。然而,杰出的日本人继续淡化他们国家的科学成就,尤其是在与美国听众交谈时。索尼的研究主管,著名的物理学家菊池诚告诉一位美国记者,日本的天才比美国少,美国是一个拥有“杰出精英”的国家。但美国也有“菊池争辩说,“智商低于正常水平”的人的“长尾巴”,解释了为什么日本更擅长大规模制造。

By the 1980s, consumer electronics had become a Japanese specialty, with Sony leading the way in launching new consumer goods, grabbing market share from American rivals. At first Japanese firms succeeded by replicating U.S. rivals’ products, manufacturing them at higher quality and lower price. Some Japanese played up the idea that they excelled at implementation, whereas America was better at innovation. “We have no Dr. Noyces or Dr. Shockleys,” one Japanese journalist wrote, though the country had begun to accumulate its share of Nobel Prize winners. Yet prominent Japanese continued to downplay their country’s scientific successes, especially when speaking to American audiences. Sony’s research director, the famed physicist Makoto Kikuchi, told an American journalist that Japan had fewer geniuses than America, a country with “outstanding elites.” But America also had “a long tail” of people “with less than normal intelligence,” Kikuchi argued, explaining why Japan was better at mass manufacturing.

美国芯片制造商坚信菊池对美国创新优势的看法是正确的,尽管相互矛盾的数据正在堆积如山。日本是“实施者”而非“创新者”这一论点的最佳证据是菊池的老板、索尼首席执行​​官森田昭夫。森田知道复制是获得二流地位和二流利润的秘诀。他驱使他的工程师不仅要制造最好的收音机和电视,而且还要完全想象新型产品。

American chipmakers clung to their belief that Kikuchi was right about America’s innovation advantage, even though contradictory data was piling up. The best evidence against the thesis that Japan was an “implementer” rather than an “innovator” was Kikuchi’s boss, Sony CEO Akio Morita. Morita knew that replication was a recipe for second-class status and second-rate profits. He drove his engineers not only to build the best radios and TVs, but to imagine new types of products entirely.

1979 年,就在安德森就美国芯片质量问题发表演讲的几个月前,索尼推出了 Walkman,这是一款彻底改变了音乐产业的便携式音乐播放器,它融合了每台设备中均采用了该公司的五个尖端集成电路。现在,世界各地的青少年可以将他们最喜欢的音乐放在口袋里,这些音乐由在硅谷首创但在日本开发的集成电路提供动力。索尼卖了全球 3.85 亿台,使 Walkman 成为历史上最受欢迎的消费设备之一。这是最纯粹的创新,它是在日本制造的。

In 1979, just months before Anderson’s presentation about quality problems in American chips, Sony introduced the Walkman, a portable music player that revolutionized the music industry, incorporating five of the company’s cutting-edge integrated circuits in each device. Now teenagers the world over could carry their favorite music in their pockets, powered by integrated circuits that had been pioneered in Silicon Valley but developed in Japan. Sony sold 385 million units worldwide, making the Walkman one of the most popular consumer devices in history. This was innovation at its purest, and it had been made in Japan.

美国曾支持日本战后转型为晶体管推销员。美国占领当局将有关晶体管发明的知识转让给日本物理学家,而华盛顿的政策制定者则确保索尼等日本公司可以轻松进入美国市场。把日本变成一个国家的目标民主资本家奏效了。现在一些美国人在问它是否运作得很好。赋予日本企业权力的战略似乎正在削弱美国的经济和技术优势。

The U.S. had supported Japan’s postwar transformation into a transistor salesman. U.S. occupation authorities transferred knowledge about the invention of the transistor to Japanese physicists, while policymakers in Washington ensured Japanese firms like Sony could easily sell into U.S. markets. The aim of turning Japan into a country of democratic capitalists had worked. Now some Americans were asking whether it had worked too well. The strategy of empowering Japanese businesses seemed to be undermining America’s economic and technological edge.

查理·斯波克(Charlie Sporck)是一位在管理 GE 生产线时被烧死的高管,他发现日本的生产力令人着迷且令人恐惧。在 Fairchild 开始芯片行业后,Sporck 离开了国家半导体公司,该公司当时是一家大型存储芯片生产商。超高效的日本竞争似乎肯定会让他破产。Sporck 以能够提高装配线工人的效率而闻名,但日本的生产力水平远远领先于他的工人所能完成的任何事情。

Charlie Sporck, the executive who’d been burned in effigy while managing a GE production line, found Japan’s productivity fascinating and frightening. After starting in the chip industry at Fairchild, Sporck left to run National Semiconductor, then a large producer of memory chips. Ultra-efficient Japanese competition seemed certain to put him out of business. Sporck had a hard-earned reputation for his ability to squeeze efficiency out of assembly line workers, but Japan’s productivity levels were far ahead of anything his workers could accomplish.

Sporck 派他的一个工头和一组流水线工人在日本花费数月时间参观半导体设施。当他们回到加利福尼亚时,Sporck 制作了一部关于他们经历的电影。他们报告说,日本工人“非常支持公司”,“工头把公司放在他的家庭之上”。日本的老板不必担心被人像烧死。这是一个“美丽的故事”,Sporck 宣称。“我们所有的员工都可以看看如何竞争很激烈。”

Sporck sent one of his foremen and a group of assembly line workers to spend several months in Japan touring semiconductor facilities. When they returned to California, Sporck made a film about their experience. They reported that Japanese workers were “amazingly pro-company” and that “the foreman put a priority to the company over his family.” Bosses in Japan didn’t have to worry about getting burned in effigy. It was a “beautiful story,” Sporck declared. “It was something for all of our employees to see how that competition is tough.”

第 16 章 “与日本交战”

CHAPTER 16 “At War with Japan”

不想假装我在公平竞争,”Advanced Micro Devices 首席执行官 Jerry Sanders 抱怨道。“我不是。” 桑德斯对打架有所了解。十八岁时,他在他长大的芝加哥南区发生争吵后差点丧命。在他的尸体在垃圾桶中被发现后,一名牧师进行了最后的仪式,尽管他在三天后奇迹般地从昏迷中苏醒。他最终在飞兆半导体找到了一份销售和营销工作,与诺伊斯、摩尔和安迪格鲁夫一起工作,然后他们离开飞兆半导体创立了英特尔。尽管他的同事大多是谦虚的工程师,但桑德斯却闪着昂贵的手表,开着一辆劳斯莱斯。他每周从他居住的南加州通勤到硅谷,因为,一位同事回忆说,他和他的妻子只有在贝莱尔才真正感到宾至如归。在创立了自己的芯片公司后,1969 年,AMD 在接下来的 30 年中的大部分时间都在与英特尔就知识产权纠纷进行法律争吵。“我无法摆脱战斗,”他对一名记者承认。

“I don’t want to pretend I’m in a fair fight,” complained Jerry Sanders, CEO of Advanced Micro Devices. “I’m not.” Sanders knew something about fights. At age eighteen, he’d almost died after a brawl on Chicago’s South Side, where he grew up. After his body was found in a garbage can, a priest administered last rites, though he miraculously emerged from a coma three days later. He eventually landed a job in sales and marketing at Fairchild Semiconductor, working alongside Noyce, Moore, and Andy Grove before they left Fairchild to found Intel. Though his colleagues were mostly modest engineers, Sanders flashed expensive watches and drove a Rolls-Royce. He commuted weekly to Silicon Valley from Southern California, where he lived, because, one colleague recalled, he and his wife only really felt at home in Bel Air. After founding his own chip firm, AMD, in 1969, he spent much of the next three decades in a legal brawl with Intel over intellectual property disputes. “I can’t walk away from a fight,” he admitted to a journalist.

“芯片行业是一个竞争异常激烈的行业,”曾领导整个亚洲芯片组装离岸外包的高管查理斯波克回忆道。“打倒他们,打倒他们,杀了他们,”Sporck 解释说,双手合十以说明他的观点。以自豪、专利和数百万美元为赌注,美国之间的争吵芯片制造商通常会变得个性化,但仍有大量增长空间。然而,日本的竞争似乎有所不同。如果 Hitachi、Fujitsu、Toshiba 和 NEC 成功,Sporck 认为,他们将把整个行业转移到太平洋彼岸。“我在 GE 专门从事电视工作,”Sporck 警告说。“你现在可以开车经过那个设施,它仍然是空的...... 我们知道危险,我们该死的,不会让这种事情发生在我们身上。” 一切都岌岌可危——工作、财富、遗产、骄傲。“我们正在与日本交战,”斯波克坚持说。“不是用枪和弹药,而是一场与技术、生产力和质量的经济战争。”

“The chip industry was an incredibly competitive industry,” remembered Charlie Sporck, the executive who’d led the offshoring of chip assembly throughout Asia. “Knock ’em down, fight ’em, kill ’em,” Sporck explained, hitting his fists together to illustrate his point. With pride, patents, and millions of dollars at stake, the brawls between U.S. chipmakers often got personal, but there was still plenty of growth to go around. Japanese competition seemed different, however. If Hitachi, Fujitsu, Toshiba, and NEC succeeded, Sporck thought, they’d move the whole industry across the Pacific. “I worked specifically on TVs at GE,” Sporck warned. “You can drive by that facility now, it’s still empty…. We knew the dangers and we damn right well weren’t gonna let that happen to us.” Everything was at stake—jobs, fortunes, legacies, pride. “We’re at war with Japan,” Sporck insisted. “Not with guns and ammunition, but an economic war with technology, productivity, and quality.”

Sporck 将硅谷的内部斗争视为公平斗争,但认为日本的 DRAM 公司受益于知识产权盗窃、受保护的市场、政府补贴和廉价资本。斯波克对间谍有意见。1981 年 11 月一个寒冷的早晨,凌晨 5 点在康涅狄格州哈特福德的一家酒店的大厅集合后,日立员工 Jun Naruse 交出了一个现金信封,并从一家名为 Glenmar 的公司的“顾问”那里收到了一张徽章承诺帮助日立获得工业机密。带着徽章,成濑进入了由飞机制造商普惠公司经营的秘密设施,并拍摄了公司最新的电脑。

Sporck saw Silicon Valley’s internal battles as fair fights, but thought Japan’s DRAM firms benefitted from intellectual property theft, protected markets, government subsidies, and cheap capital. Sporck had a point about the spies. After a 5 a.m. rendezvous in the lobby of a Hartford, Connecticut, hotel on a cold November morning in 1981, Hitachi employee Jun Naruse handed over an envelope of cash and received in exchange a badge from a “consultant” at a company called Glenmar that promised to help Hitachi obtain industrial secrets. With the badge, Naruse gained entrance to a secret facility run by aircraft maker Pratt & Whitney and photographed the company’s newest computer.

拍摄结束后,成濑在西海岸的同事林健二给 Glenmar 写了一封信,提议签订一份“咨询服务合同”。日立的高级管理人员授权向 Glenmar 支付 50 万美元以继续这种关系。但 Glenmar 是一个幌子公司。它的雇员是联邦调查局特工。“看来是日立进了圈套,”该公司发言人在日立的员工被捕并登上《纽约时报》商业版头版后羞怯地承认。

After the photo shoot, Naruse’s colleague on the West Coast, Kenji Hayashi, sent a letter to Glenmar proposing a “consultation service contract.” Hitachi’s senior executives authorized half a million dollars in payments to Glenmar to continue the relationship. But Glenmar was a front company; its employees were FBI agents. “It seems that Hitachi stepped into the trap,” the company’s spokesman sheepishly admitted, after Hitachi’s employees were arrested and the story made the front page of the business section of the New York Times.

日立并不孤单。三菱电机面临类似指控。对日本间谍活动和双重交易的指控不仅在半导体和计算机领域上演。东芝是一家日本工业集团,在 1980 年代中期已成为世界领先的 DRAM 生产商,多年来一直在反对声称——事实证明——该公司出售了帮助他们建造的苏联机器更安静的潜艇。东芝的苏联潜艇交易与该公司的半导体业务之间没有直接联系,但许多美国人认为潜艇案是日本人的进一步证据肮脏的交易。记录在案的日本非法工业间谍案件数量很少。但这是否表明窃取机密在日本的成功中只起到了很小的作用,还是日本公司擅长间谍活动的证据?

Hitachi wasn’t alone. Mitsubishi Electric faced similar charges. It wasn’t only in semiconductors and computers that accusations of Japanese espionage and double-dealing swirled. Toshiba, the Japanese industrial conglomerate that by the mid-1980s was a world-leading DRAM producer, spent years fighting claims—true, it turned out—that the company sold the Soviets machinery that helped them build quieter submarines. There was no direct link between Toshiba’s Soviet submarine deal and the company’s semiconductor business, but many Americans saw the submarine case as further evidence of Japanese dirty dealing. The number of documented cases of illegal Japanese industrial espionage was low. But was this a sign that stealing secrets played only a small role in Japan’s success, or evidence that Japanese firms were skilled at spycraft?

潜入竞争对手的设施是非法的,但密切关注竞争对手是硅谷的正常做法。因此,也有人指责竞争对手窃取员工、创意和知识产权。毕竟,美国的芯片制造商一直在互相起诉。例如,Fairchild 和德州仪器之间经过十年的诉讼才解决了诺伊斯或基尔比是否发明了集成电路的问题。芯片公司也经常挖走竞争对手的明星工程师,不仅希望获得经验丰富的工人,而且还希望获得有关竞争对手生产流程的知识。Noyce 和 Moore 离开 Shockley Semiconductor 创立了 Fairchild,然后离开 Fairchild 创立了 Intel,在那里他们雇佣了包括 Andy Grove 在内的数十名 Fairchild 员工。仙童在决定不可能赢得针对创建芯片行业的天才的诉讼之前考虑起诉。追踪和模仿竞争对手是硅谷商业模式的关键。日本的战略有什么不同吗?

Sneaking into rivals’ facilities was illegal but keeping tabs on competitors was normal practice in Silicon Valley. So, too, was accusing rivals of pilfering employees, ideas, and intellectual property. America’s chipmakers were constantly suing each other, after all. It took a decade of litigation between Fairchild and Texas Instruments to resolve the question of whether Noyce or Kilby had invented the integrated circuit, for example. Chip firms regularly poached rivals’ star engineers, too, hoping not only to acquire experienced workers but also knowledge about their competitors’ production processes. Noyce and Moore had left Shockley Semiconductor to found Fairchild, then left Fairchild to found Intel, where they hired dozens of Fairchild employees, including Andy Grove. Fairchild considered suing before deciding that it was unlikely to win a lawsuit against the geniuses who had built the chip industry. Tracking and emulating rivals was key to Silicon Valley’s business model. Was Japan’s strategy any different?

Sporck 和 Sanders 指出,日本公司也受益于受保护的国内市场。日本公司可以向美国出售产品,但硅谷难以在日本赢得市场份额。直到 1974 年,日本实施配额限制美国公司可以在那里销售的芯片数量。即使取消了这些配额,日本公司仍然很少从硅谷购买芯片,尽管日本消耗了全球四分之一的半导体,索尼等公司将这些半导体插入到全球销售的电视和录像机中。日本的一些大型芯片消费者,如日本国家电信垄断企业 NTT,几乎完全从日本供应商那里购买。这是表面上是一个商业决策,但 NTT 是政府所有的,所以政治可能发挥了作用。硅谷的日本的低市场份额使美国公司损失了数十亿美元的销售额。

Sporck and Sanders pointed out that Japanese firms benefitted from a protected domestic market, too. Japanese firms could sell to the U.S., but Silicon Valley struggled to win market share in Japan. Until 1974, Japan imposed quotas limiting the number of chips U.S. firms could sell there. Even after these quotas were lifted, Japanese companies still bought few chips from Silicon Valley, even though Japan consumed a quarter of the world’s semiconductors, which companies like Sony plugged into TVs and VCRs that were sold worldwide. Some big Japanese chip consumers such as NTT, Japan’s national telecom monopoly, bought almost exclusively from Japanese suppliers. This was ostensibly a business decision, but NTT was government-owned, so politics likely played a role. Silicon Valley’s low market share in Japan cost American companies billions of dollars in sales.

日本政府也为其芯片制造商提供补贴。与美国的反垄断法不鼓励芯片公司合作不同,日本政府推动公司合作,于 1976 年成立了一个名为 VLSI Program 的研究联盟,政府资助约为预算的一半。美国的芯片制造商将此作为日本不公平竞争的证据,尽管 VLSI 计划每年用于研发的 7200 万美元与德州仪器的研发预算大致相同,但低于摩托罗拉。此外,美国政府本身也深入参与了支持半导体的工作,尽管华盛顿的资金来自美国国防部高级研究计划局(DARPA)的赠款,该机构投资于投机性技术,并在为芯片制造创新提供资金方面发挥了关键作用。

Japan’s government subsidized its chipmakers, too. Unlike in the U.S., where antitrust law discouraged chip firms from collaborating, the Japanese government pushed companies to work together, launching a research consortium called the VLSI Program in 1976 with the government funding around half the budget. America’s chipmakers cited this as evidence of unfair Japanese competition, though the $72 million the VLSI Program spent annually on R&D was about the same as Texas Instruments’ R&D budget, and less than Motorola’s. Moreover, the U.S. government was itself deeply involved in supporting semiconductors, though Washington’s funding took the form of grants from DARPA, the Pentagon unit that invests in speculative technologies and has played a crucial role in funding chipmaking innovation.

Jerry Sanders 认为硅谷最大的劣势在于其高昂的资本成本。日本人“为资本支付 6%,也许是 7%。我付18% 在美好的一天,”他抱怨道。建造先进的制造设施非常昂贵,因此信贷成本非常重要。下一代芯片大约每两年出现一次,需要新设施和新机器。在 1980 年代,随着美联储寻求对抗通货膨胀,美国利率达到 21.5%。

Jerry Sanders saw Silicon Valley’s biggest disadvantage as its high cost of capital. The Japanese “pay 6 percent, maybe 7 percent, for capital. I pay 18 percent on a good day,” he complained. Building advanced manufacturing facilities was brutally expensive, so the cost of credit was hugely important. A next-generation chip emerged roughly once every two years, requiring new facilities and new machinery. In the 1980s, U.S. interest rates reached 21.5 percent as the Federal Reserve sought to fight inflation.

相比之下,日本 DRAM 公司获得了便宜得多的资本。日立和三菱等芯片制造商是大型企业集团的一部分,它们与提供大量长期贷款的银行有着密切的联系。即使在日本公司无利可图的情况下,他们的银行也会通过在美国放贷机构提供信贷之后很久才提供信贷来维持他们的生存。逼他们破产。日本社会在结构上倾向于产生大量储蓄,因为战后的婴儿潮和向独生子女家庭的迅速转变造成了大量专注于为退休储蓄的中年家庭。日本的社会保障不足净额为储蓄提供了进一步的动力。与此同时,对股市和其他投资的严格限制让人们别无选择,只能将储蓄存入银行账户。结果,银行存款充裕,以低利率发放贷款,因为他们手头有很多现金。日本公司的债务比美国同行多,但尽管如此支付较低的利率来借款。

By contrast, Japanese DRAM firms got access to far cheaper capital. Chipmakers like Hitachi and Mitsubishi were part of vast conglomerates with close links to banks that provided large, long-term loans. Even when Japanese companies were unprofitable, their banks kept them afloat by extending credit long after American lenders would have driven them to bankruptcy. Japanese society was structurally geared to produce massive savings, because its postwar baby boom and rapid shift to one-child households created a glut of middle-aged families focused on saving for retirement. Japan’s skimpy social safety net provided a further incentive for saving. Meanwhile, tight restrictions on stock markets and other investments left people with little choice but to stuff savings in bank accounts. As a result, banks were flush with deposits, extending loans at low rates because they had so much cash on hand. Japanese companies had more debt than American peers but nevertheless paid lower rates to borrow.

凭借这笔廉价资金,日本企业展开了一场无情的市场份额争夺战。尽管一些美国分析家描绘了合作的形象,但东芝、富士通和其他公司在相互竞争中同样无情。然而,由于可以获得几乎无限的银行贷款,他们在等待竞争对手破产时可能会蒙受损失。在 1980 年代初期,日本公司在生产设备上的投资比美国竞争对手高出 60%,尽管该行业的每个人都面临着同样残酷的竞争,几乎没有人能获得多少利润。日本芯片制造商不断投资和生产,抢占越来越多的市场份额。正因为如此,在 64K DRAM 芯片推出五年后,十年前开创 DRAM 芯片的英特尔公司只剩下全球DRAM市场1.7%的份额,而日本竞争对手的市场份额飙升。

With this cheap capital, Japanese firms launched a relentless struggle for market share. Toshiba, Fujitsu, and others were just as ruthless in competing with each other, despite the cooperative image painted by some American analysts. Yet with practically unlimited bank loans available, they could sustain losses as they waited for competitors to go bankrupt. In the early 1980s, Japanese firms invested 60 percent more than their U.S. rivals in production equipment, even though everyone in the industry faced the same cutthroat competition, with hardly anyone making much profit. Japanese chipmakers kept investing and producing, grabbing more and more market share. Because of this, five years after the 64K DRAM chip was introduced, Intel—the company that had pioneered DRAM chips a decade earlier—was left with only 1.7 percent of the global DRAM market, while Japanese competitors’ market share soared.

随着硅谷的退出,日本公司的 DRAM 产量翻了一番。1984 年,日立在其半导体业务上的资本支出为 800 亿日元,而十年前为 15 亿日元。在东芝,支出从 30 亿增长到 750 亿;在NEC,从35亿增加到1100亿。1985 年,日本公司将世界资本支出的 46% 用于半导体,而美国为 35%。到 1990 年,数字更加不平衡,日本公司占世界芯片制造设施和设备投资的一半。日本的 CEO 们不断建设新的设施,只要他们的银行还在很高兴买单。

Japan’s firms doubled down on DRAM production as Silicon Valley was pushed out. In 1984, Hitachi spent 80 billion yen on capital expenditure for its semiconductor business, compared to 1.5 billion a decade earlier. At Toshiba, spending grew from 3 billion to 75 billion; at NEC, from 3.5 billion to 110 billion. In 1985, Japanese firms spent 46 percent of the world’s capital expenditure on semiconductors, compared to America’s 35 percent. By 1990, the figures were even more lopsided, with Japanese firms accounting for half the world’s investment in chipmaking facilities and equipment. Japan’s CEOs kept building new facilities so long as their banks were happy to foot the bill.

日本芯片制造商认为,这一切都不公平。美国的半导体公司从政府那里得到了很多帮助,尤其是通过国防合同。无论如何,美国的芯片消费者,喜欢惠普有确凿的证据表明日本芯片的质量更好。因此,在 1980 年代,日本在 DRAM 芯片中的市场份额每年都在增长,而美国竞争对手则以牺牲为代价。不管美国芯片制造商的世界末日预言如何,日本的半导体激增似乎势不可挡。很快,整个硅谷都会死去,就像十几岁的杰里桑德斯在南边的垃圾桶里一样。

The Japanese chipmakers argued that none of this was unfair. America’s semiconductor firms got plenty of help from the government, especially via defense contracts. Anyway, American consumers of chips, like HP, had hard evidence that Japanese chips were simply better quality. So Japan’s market share in DRAM chips grew every year during the 1980s, at the expense of American rivals. Japan’s semiconductor surge seemed unstoppable, no matter the apocalyptic predictions of American chipmakers. Soon all of Silicon Valley would be left for dead, like teenage Jerry Sanders in a South Side garbage can.

第十七章 “航运垃圾”

CHAPTER 17 “Shipping Junk”

随着日本巨头横扫美国的高科技产业,苦苦挣扎的不仅是生产 DRAM 芯片的公司。他们的许多供应商也这样做了。1981 年,GCA 公司被庆祝为美国的“最热门的高科技公司”,通过销售使摩尔定律成为可能的设备而迅速发展。自从物理学家 Jay Lathrop 首次将他的显微镜倒置以照射光刻胶化学品并在半导体晶圆上“打印”图案以来的二十年间,光刻工艺变得更加复杂。那些日子早已一去不复返了鲍勃·诺伊斯(Bob Noyce)开着他的老爷车在加利福尼亚州的 101 号高速公路上行驶,为飞兆半导体的临时光刻设备寻找电影摄影机镜头。现在光刻是一项大生意,在 1980 年代初期,GCA 处于领先地位。

As the Japanese juggernaut tore through America’s high-tech industry, it wasn’t only companies producing DRAM chips that struggled. Many of their suppliers did, too. In 1981, GCA Corporation was being celebrated as one of America’s “hottest high-technology corporations,” growing rapidly by selling equipment that made possible Moore’s Law. In the two decades since physicist Jay Lathrop had first turned his microscope upside down to shine light on photoresist chemicals and “print” patterns on semiconductor wafers, the process of photolithography had become vastly more complicated. Long gone were the days of Bob Noyce driving up and down California’s Highway 101 in his old jalopy in search of movie camera lenses for Fairchild’s makeshift photolithography equipment. Now lithography was big business, and at the start of the 1980s, GCA was at the top.

尽管光刻技术比 Jay Lathrop 的倒置显微镜时代要精确得多,但原理保持不变。光线透过掩模和透镜照射,将聚焦的形状投射到覆盖有光刻胶化学物质的硅晶片上。在光线照射的地方,化学物质与光线发生反应,使它们被冲走,露出硅晶片顶部的微观凹痕。在这些孔中添加了新材料,在硅上构建电路。专用化学品蚀刻掉光刻胶,留下完美成型的形状背后。制造集成电路通常需要五、十或二十次光刻、沉积、蚀刻和抛光迭代,其结果就像几何婚礼蛋糕一样分层。随着晶体管的小型化,光刻工艺的每个部分——从化学品到透镜再到将硅晶片与光源完美对齐的激光器——变得更加困难。

Though photolithography had become far more precise than in the days of Jay Lathrop’s upside-down microscope, the principles remained the same. A light shined through masks and lenses, projecting focused shapes onto a silicon wafer covered with photoresist chemicals. Where light struck, the chemicals reacted with the light, allowing them to be washed away, exposing microscopic indentations on top of the silicon wafer. New materials were added in these holes, building circuits on the silicon. Specialized chemicals etched away the photoresist, leaving behind perfectly formed shapes. It often took five, ten, or twenty iterations of lithography, deposition, etching, and polishing to fabricate an integrated circuit, with the result layered like a geometric wedding cake. As transistors were miniaturized, each part of the lithography process—from the chemicals to the lenses to the lasers that perfectly aligned the silicon wafers with the light source—became even more difficult.

世界领先的镜头制造商是德国的卡尔蔡司和日本的尼康,尽管美国也有一些专业的镜头制造商。Perkin Elmer 是康涅狄格州诺沃克的一家小型制造商,在二战期间为美军制造了炸弹瞄准具,并为冷战时期的卫星和间谍飞机制造了镜头。该公司意识到这项技术可用于半导体光刻技术,并开发了一种芯片扫描仪,它可以以几乎完美的精度对齐硅晶片和光刻光源,如果光线要完全按照预期照射到硅片上,这一点至关重要。这台机器像复印机一样将光线穿过晶片,将覆盖有光刻胶的晶片曝光,就好像它被涂上了光线一样。Perkin Elmer 的扫描仪可以制造出宽度接近一微米(百万分之一米)的芯片。

The world’s leading lens makers were Germany’s Carl Zeiss and Japan’s Nikon, though the U.S. had a few specialized lens makers, too. Perkin Elmer, a small manufacturer in Norwalk, Connecticut, had made bombsights for the U.S. military during World War II and lenses for Cold War satellites and spy planes. The company realized this technology could be used in semiconductor lithography and developed a chip scanner that could align a silicon wafer and a lithographic light source with almost perfect precision, which was crucial if the light was to hit the silicon exactly as intended. The machine moved the light across the wafer like a copy machine, exposing the photoresist-covered wafer as if it were being painted with lines of light. Perkin Elmer’s scanner could create chips with features approaching one micron—a millionth of a meter—in width.

Perkin Elmer 的扫描仪在 1970 年代后期主导了光刻市场,但到了 1980 年代,它已被 GCA 取代,GCA 是一家由空军军官出身的地球物理学家米尔特格林伯格领导的公司,他是一位雄心勃勃、固执、口齿不清的天才。格林伯格和一位空军伙伴在二战后用洛克菲勒家族的种子资金创立了 GCA。作为一名军事气象学家,格林伯格将他对大气的知识和他与空军的联系运用到作为国防承包商的工作中,生产出像高空气球这样的设备,这些设备可以进行测量并采取苏联的照片。

Perkin Elmer’s scanner dominated the lithography market in the late 1970s, but by the 1980s, it had been displaced by GCA, a company led by an Air Force officer–turned-geophysicist named Milt Greenberg, an ambitious, stubborn, foul-mouthed genius. Greenberg and an Air Force buddy founded GCA after World War II with seed capital from the Rockefellers. Trained as a military meteorologist, Greenberg had parlayed his knowledge of the atmosphere and his Air Force connections into work as a defense contractor, producing devices like high-altitude balloons that made measurements and took photographs of the Soviet Union.

格林伯格的野心很快就飞得更高了。半导体行业的增长表明,真正的资金来自大众市场,而不是专门的军事合同。格林伯格认为他的公司用于军事侦察的高科技光学系统可以部署在民用芯片上。在 1970 年代后期的一次行业会议上,GCA 为芯片制造商宣传其系统,德州仪器Morris Chang 走到 GCA 的展台前,开始查看公司的设备,询问公司的设备是否可以逐步移动,而不是扫描整个晶圆的光,从而暴露硅晶圆上的每个芯片。这样的“步进器”将比现有的扫描仪准确得多。虽然从未设计过步进器,但 GCA 的工程师相信他们可以制造出一个,提供更高分辨率的成像,从而提供更小的晶体管。

Greenberg’s ambitions soon flew even higher. The growth in the semiconductor industry showed that the real money was in the mass market, not in specialized military contracts. Greenberg thought his company’s high-tech optical systems—useful for military reconnaissance—could be deployed on civilian chips. At an industry conference in the late 1970s, where GCA was advertising its systems for chipmakers, Texas Instruments’ Morris Chang walked up to the GCA booth, started looking at the company’s equipment, and inquired whether, rather than scanning light across the length of a wafer, the firm’s equipment could move step-by-step, exposing each chip on the silicon wafer. Such a “stepper” would be far more accurate than the existing scanners. Though a stepper had never been devised, GCA’s engineers believed they could create one, providing higher-resolution imaging and thus smaller transistors.

几年后的 1978 年,GCA 推出了它的第一个步进器。销售订单开始滚滚而来。在步进电机之前,GCA 的军事合同年收入从未超过 5000 万美元,但现在它垄断了一台非常有价值的机器。收入很快达到 3 亿美元,公司的股价飙升。

Several years later, in 1978, GCA introduced its first stepper. Sales orders began rolling in. Before the stepper, GCA had never made more than $50 million a year in revenue on its military contracts, but now it had a monopoly on an extraordinarily valuable machine. Revenue soon hit $300 million and the company’s stock price surged.

然而,随着日本芯片产业的崛起,GCA 开始失去优势。首席执行官格林伯格把自己想象成一个商业巨头,但他花在经营企业上的时间更少,而更多地与政客打成一片。他在一座新的大型制造设施上破土动工,押注 1980 年代初期的半导体繁荣将无限期地持续下去。成本失控。库存管理不善。一名员工无意中发现了被遗忘在壁橱里的价值 100 万美元的精密镜片。高管们用公司信用卡购买克尔维特的故事流传开来。格林伯格的一位创始合伙人承认,公司花钱就像“喝醉的水手。”

As Japan’s chip industry rose, however, GCA began to lose its edge. Greenberg, the CEO, imagined himself as a business titan, but he spent less time running the business and more hobnobbing with politicians. He broke ground on a major new manufacturing facility, betting that the early 1980s semiconductor boom would continue indefinitely. Costs spun out of control. Inventory was wildly mismanaged. One employee stumbled onto a million dollars’ worth of precision lenses sitting forgotten in a closet. Stories circulated of executives buying Corvettes on company credit cards. One of Greenberg’s founding partners admitted that the company was spending money like a “drunken sailor.”

该公司的过度行为时机不当。半导体行业的周期性一直很凶猛,需求旺盛时行业暴涨,需求不旺时则回落。不需要一位火箭科学家(而 GCA 有少数员工)就发现,在 1980 年代初期的繁荣之后,最终会出现衰退。格林伯格选择不听。“他不想从营销部门听到'将会出现低迷',”一名员工回忆道。于是公司进入了1980年代中期半导体暴跌严重过度。1984 年至 1986 年间,光刻设备的全球销售额下降了 40%。GCA 的收入下降了三分之二以上。“如果我们的员工中有一位称职的经济学家,我们可能已经预测到了,”一名员工回忆道。“但我们没有。我们有米尔特。”

The firm’s excesses were poorly timed. The semiconductor industry had always been ferociously cyclical, with the industry skyrocketing upward when demand was strong, and slumping back when it was not. It didn’t take a rocket scientist—and GCA had a handful on staff—to figure out that after the boom of the early 1980s, a downturn would eventually follow. Greenberg chose not to listen. “He didn’t want to hear from the marketing department that ‘there’s going to be a downturn,’ ” one employee remembered. So the company entered the mid-1980s semiconductor slump heavily overextended. Global sales of lithography equipment fell by 40 percent between 1984 and 1986. GCA’s revenue fell by over two-thirds. “If we had a competent economist on staff, we might have predicted it,” one employee remembered. “But we didn’t. We had Milt.”

就在市场暴跌之际,GCA 失去了作为唯一一家建立步进器的公司的地位。日本尼康最初是 GCA 的合作伙伴,为其步进提供精密镜头。但格林伯格决定退出尼康,收购了他自己的镜头制造商,总部位于纽约的 Tropel,该公司为 U2 间谍飞机生产镜头,但难以生产 GCA 所需的高质量镜头数量。与此同时,GCA 的客户服务萎缩了。一位分析师回忆说,该公司的态度是“购买我们制造的产品,不要打扰我们”。公司自己的员工承认“客户受够了。” 这是垄断者的态度——但 GCA 不再是垄断者。在格林伯格停止购买尼康镜头后,这家日本公司决定制造自己的步进器。它从 GCA 购买了一台机器并对其进行了逆向工程。很快,尼康的市场份额就超过了 GCA。

Just as the market slumped, GCA lost its position as the only company building steppers. Japan’s Nikon had initially been a partner of GCA, providing the precision lenses for its stepper. But Greenberg had decided to cut Nikon out, buying his own lens maker, New York−based Tropel, which made lenses for the U2 spy planes but which struggled to produce the number of high-quality lenses GCA needed. Meanwhile, GCA’s customer service atrophied. The company’s attitude, one analyst recounted, was “buy what we build and don’t bother us.” The company’s own employees admitted that “customers got fed up.” This was the attitude of a monopolist—but GCA was no longer a monopoly. After Greenberg stopped buying Nikon lenses, the Japanese company decided to make its own stepper. It acquired a machine from GCA and reverse engineered it. Soon Nikon had more market share than GCA.

许多美国人将 GCA 失去光刻领导地位归咎于日本的工业补贴。的确,日本的 VLSI 计划提振了该国的 DRAM 芯片生产商,也帮助了像尼康这样的设备供应商。随着美国和日本公司相互指责政府帮助不公平,商业关系变得风雨飘摇。但 GCA 员工承认,虽然他们的技术是世界一流的,但公司在大规模生产方面遇到了困难。精密制造是必不可少的,因为现在光刻技术非常精确,以至于滚过的雷暴可能会改变气压——从而改变光线折射的角度——足以扭曲刻在芯片上的图像。每年制造数百台步进电机需要将激光重点放在制造和质量控制上。但 GCA 的领导者专注于其他地方。

Many Americans blamed Japan’s industrial subsidies for GCA’s loss of lithography leadership. It was true that Japan’s VLSI program, which boosted the country’s producers of DRAM chips, also helped equipment suppliers like Nikon. As U.S. and Japanese firms traded accusations of unfair government help, commercial relations grew stormy. But GCA employees admitted that, though their technology was world class, the company struggled with mass production. Precision manufacturing was essential, since lithography was now so exact that a thunderstorm rolling through could change air pressure—and thus the angle at which light refracted—enough to distort the images carved on chips. Building hundreds of steppers a year required a laser focus on manufacturing and quality control. But GCA’s leaders were focused elsewhere.

人们普遍将 GCA 的衰落解释为日本崛起和美国衰落的寓言。一些分析师看到了更广泛的证据制造业衰退始于钢铁,然后是汽车,现在正在蔓延到高科技行业。1987 年,诺贝尔奖获得者、麻省理工学院经济学家罗伯特·索洛(Robert Solow)率先研究了生产力和经济增长,他认为芯片行业存在“不稳定的结构”,员工在公司之间跳槽,而公司则拒绝投资于员工。著名经济学家罗伯特·赖希(Robert Reich)感叹硅谷的“纸上谈兵”,他认为这种做法过于注重追求声望和财富,而不是技术进步。他宣称,在美国的大学里,“科学和工程项目正在失败。”

It was popular to interpret the decline of GCA as an allegory about Japan’s rise and America’s fall. Some analysts saw evidence of a broader manufacturing decay that started in steel, then afflicted cars, and was now spreading to high-tech industries. In 1987, Nobel Prize−winning MIT economist Robert Solow, who pioneered the study of productivity and economic growth, argued that the chip industry suffered from an “unstable structure,” with employees job hopping between firms and companies declining to invest in their workers. Prominent economist Robert Reich lamented the “paper entrepreneurialism” in Silicon Valley, which he thought focused too much on the search for prestige and affluence rather than technical advances. At American universities, he declared, “science and engineering programs are foundering.”

美国芯片制造商的 DRAM 灾难与 GCA 市场份额的下滑有一定关系。与硅谷竞争的日本 DRAM 公司更愿意从日本工具制造商那里购买,以牺牲 GCA 为代价使尼康受益。然而,GCA 的大部分问题都是由不可靠的设备和糟糕的客户服务造成的。学者们设计了详尽的理论来解释日本的大型企业集团如何比美国的小型初创企业更擅长制造。但世俗的现实是 GCA 没有倾听客户的意见,而尼康却做到了。与 GCA 互动的芯片公司发现了它“傲慢”和“没有反应”。没有人这么说它的日本竞争对手。

American chipmakers’ DRAM disaster was somewhat related to GCA’s collapsing market share. The Japanese DRAM firms that were outcompeting Silicon Valley preferred to buy from Japanese toolmakers, benefitting Nikon at the expense of GCA. However, most of GCA’s problems were homegrown, driven by unreliable equipment and bad customer service. Academics devised elaborate theories to explain how Japan’s huge conglomerates were better at manufacturing than America’s small startups. But the mundane reality was that GCA didn’t listen to its customers, while Nikon did. Chip firms that interacted with GCA found it “arrogant” and “not responsive.” No one said that about its Japanese rivals.

因此,到 1980 年代中期,尼康的系统比 GCA 的系统要好得多——即使天空晴朗。尼康的机器产生了明显更好的产量,而且发生故障的频率要低得多。例如,在 IBM 过渡到尼康步进器之前,它希望它使用的每台机器都能工作 75 小时,然后才需要停机进行调整或维修。尼康的客户平均持续使用时间的十倍。

By the mid-1980s, therefore, Nikon’s systems were far better than GCA’s—even when the skies were sunny. Nikon’s machines produced meaningfully better yields and broke down far less often. Before IBM transitioned to Nikon steppers, it hoped each machine it used would work seventy-five hours before needing downtime for adjustments or repairs, for example. Nikon’s customers averaged ten times that duration of continuous use.

GCA 的首席执行官格林伯格永远无法弄清楚如何修复公司。直到他被赶下台的那天,他并没有意识到他的公司有多少问题是内部的。当他环游世界进行销售访问时,头等舱喝着血腥玛丽,客户认为该公司正在“运送垃圾”。员工抱怨格林伯格与华尔街息息相关,既关注股票价格,也关注商业模式。为了获得年终数据,该公司将与客户勾结,在 12 月运送一个装有用户手册的空板条箱,然后在第二年自行交付机器。然而,这无法掩盖公司失去的市场份额。1978 年,以 GCA 为首的美国公司控制了全球 85% 的半导体光刻设备市场。十年后,这一数字下降到 50%。GCA有没有扭转局面的计划。

Greenberg, GCA’s CEO, could never figure out how to fix the company. Up to the day he was ousted, he didn’t realize just how many of his company’s problems were internal. As he flew around the world on sales visits, drinking a Bloody Mary in first class, customers thought the firm was “shipping junk.” Employees complained that Greenberg was in hock to Wall Street, focused as much on the stock price as on the business model. To make end-of-year numbers, the company would collude with customers, shipping an empty crate with a user’s manual in December before delivering the machines themselves the subsequent year. However, it was impossible to cover up the company’s loss of market share. U.S. firms, with GCA as the leader, controlled 85 percent of the global market for semiconductor lithography equipment in 1978. A decade later this figure had dropped to 50 percent. GCA had no plan to turn things around.

格林伯格本人将批评的矛头对准了公司的员工。“他会使用令人难以置信的四个字母的单词,”一名下属回忆道。另一个人回忆起禁止高跟鞋的决定,格林伯格认为这毁了公司的地毯。随着紧张局势的加剧,接待员与同事制定了一个代码,打开天花板灯以表示格林伯格在大楼内,并在他离开时将其关闭。当他出去的时候,每个人都可以轻松一点。但这并不能阻止美国的光刻机领导者冲向危机。

Greenberg himself aimed criticism at the company’s employees. “He would use unbelievable four-letter words,” one subordinate remembered. Another recalled a decision to ban high-heeled shoes, which Greenberg thought ruined the company’s carpets. As tension grew, the receptionist developed a code with fellow employees, turning on a ceiling light to denote that Greenberg was in the building, and turning it off when he left. Everyone could breathe a bit easier when he was out. But this couldn’t stop America’s lithography leader from hurtling toward crisis.

第十八章 1980 年代的原油

CHAPTER 18 The Crude Oil of the 1980s

在帕洛阿尔托一个寒冷的春夜,鲍勃·诺伊斯、杰里·桑德斯和查理·斯波克相遇在倾斜的宝塔式屋顶下。Ming's Chinese Restaurant 是硅谷午餐圈的主食。但美国的科技巨头们并没有因为著名的中国鸡肉沙拉而在 Ming's。Noyce、Sanders 和 Sporck 都在 Fairchild 开始了他们的职业生涯:Noyce 是技术远见者;营销表演者桑德斯;制造业老板斯波克对他的员工咆哮着要建造得更快、更便宜、更好。十年后,他们成为美国三大芯片制造商的首席执行官,成为竞争对手。但随着日本市场份额的增长,他们决定是时候再次联合起来了。美国半导体产业的未来岌岌可危。他们挤在 Ming's 私人餐厅的一张桌子旁,想出了一个新的策略来拯救它。在无视政府十年后,他们转向华盛顿寻求帮助。

On a chilly spring evening in Palo Alto, Bob Noyce, Jerry Sanders, and Charlie Sporck met under a sloping, pagoda-style roof. Ming’s Chinese Restaurant was a staple of the Silicon Valley lunch circuit. But America’s tech titans weren’t at Ming’s for its famous Chinese chicken salad. Noyce, Sanders, and Sporck had all started their careers at Fairchild: Noyce the technological visionary; Sanders the marketing showman; Sporck the manufacturing boss barking at his employees to build faster, cheaper, better. A decade later they’d become competitors as CEOs of three of America’s biggest chipmakers. But as Japan’s market share grew, they decided it was time to band together again. At stake was the future of America’s semiconductor industry. Huddled over a table in a private dining room at Ming’s, they devised a new strategy to save it. After a decade of ignoring the government, they were turning to Washington for help.

半导体是“1980 年代的原油,”杰里桑德斯宣称,“控制原油的人将控制电子工业。” 作为美国最大的芯片制造商之一 AMD 的首席执行官,桑德斯有很多自私的理由将他的主要产品描述为具有战略意义的关键产品。但他错了吗?在整个 1980 年代,美国的计算机产业迅速扩张,因为个人电脑变得足够小且价格便宜,足以供个人家庭或办公室使用。每个企业都开始依赖他们。没有集成电路,计算机就无法工作。到 1980 年代,飞机、汽车、摄像机、微波炉或 Sony Walkman 也不行。现在每个美国人的房子和汽车里都有半导体;许多人每天使用几十个芯片。就像石油一样,没有它们就无法生存。这不是让他们“战略化”了吗?美国不应该担心日本正在成为“沙特阿拉伯的半导体”?

Semiconductors are the “crude oil of the 1980s,” Jerry Sanders declared, “and the people who control the crude oil will control the electronics industry.” As CEO of AMD, one of America’s biggest chipmakers, Sanders had plenty of self-interested reasons to describe his main product as strategically crucial. But was he wrong? Throughout the 1980s, America’s computer industry expanded rapidly, as PCs were made small enough and cheap enough for an individual home or office. Every business was coming to rely on them. Computers couldn’t work without integrated circuits. Nor, by the 1980s, could planes, automobiles, camcorders, microwaves, or the Sony Walkman. Every American now had semiconductors in their houses and cars; many used dozens of chips daily. Like oil, they were impossible to live without. Didn’t this make them “strategic”? Shouldn’t America be worried Japan was becoming “the Saudi Arabia of semiconductors”?

1973 年和 1979 年的石油禁运向许多美国人证明了依赖外国生产的风险。当阿拉伯政府削减石油出口以惩罚美国支持以色列时,美国经济陷入了痛苦的衰退。十年的滞胀和政治危机接踵而至。美国的外交政策专注于波斯湾并确保其石油供应。吉米卡特总统宣布该地区是“美利坚合众国的切身利益”之一。罗纳德·里根派美国海军护送油轮进出海湾。乔治·H·W·布什与伊拉克开战的部分原因是为了解放科威特的油田。当美国说石油是一种“战略”商品时,它以军事力量支持这一说法。

The oil embargoes of 1973 and 1979 had demonstrated to many Americans the risks of relying on foreign production. When Arab governments cut oil exports to punish America for supporting Israel, the U.S. economy plunged into a painful recession. A decade of stagflation and political crises followed. American foreign policy fixated on the Persian Gulf and securing its oil supplies. President Jimmy Carter declared the region one of “the vital interests of the United States of America.” Ronald Reagan deployed the U.S. Navy to escort oil tankers in and out of the Gulf. George H. W. Bush went to war with Iraq in part to liberate Kuwait’s oil fields. When America said that oil was a “strategic” commodity, it backed the claim with military force.

桑德斯并没有要求美国派遣海军穿越半个地球来确保硅的供应。但政府难道不应该想办法帮助陷入困境的半导体公司吗?在 1970 年代,硅谷公司已经忘记了政府,因为他们用民用计算机和计算器市场取代了国防合同。1980 年代,他们羞怯地爬回华盛顿。在 Ming's 吃完晚饭后,Sanders、Noyce 和 Sporck 与其他 CEO 一起创建了半导体行业协会,游说华盛顿支持该行业。

Sanders wasn’t asking for the U.S. to send the Navy halfway across the world to secure supplies of silicon. But shouldn’t the government find a way to help its struggling semiconductor firms? In the 1970s, Silicon Valley firms had forgotten about the government as they replaced defense contracts with civilian computer and calculator markets. In the 1980s, they crawled sheepishly back to Washington. After their dinner at Ming’s, Sanders, Noyce, and Sporck joined other CEOs to create the Semiconductor Industry Association to lobby Washington to support the industry.

当杰里桑德斯将芯片描述为“原油”时,五角大楼完全明白他的意思。事实上,芯片比石油更具战略意义。五角大楼官员知道半导体对美国军事霸权的重要性。自 1970 年代中期以来,利用半导体技术“抵消”苏联在冷战中的​​常规优势一直是美国的战略,当时鲍勃·诺伊斯 (Bob Noyce) 的歌唱搭档比尔佩里负责五角大楼的研究和工程部门。美国国防公司被指示在他们最新的飞机、坦克和火箭上装上尽可能多的芯片,以实现更好的制导、通信和指挥控制。在产生军事力量方面,该战略比比尔·佩里以外的任何人都认为可能的效果更好。

When Jerry Sanders described chips as “crude oil,” the Pentagon knew exactly what he meant. In fact, chips were even more strategic than petroleum. Pentagon officials knew just how important semiconductors were to American military primacy. Using semiconductor technology to “offset” the Soviet conventional advantage in the Cold War had been American strategy since the mid-1970s, when Bob Noyce’s singing partner Bill Perry ran the Pentagon’s research and engineering division. American defense firms had been instructed to pack their newest planes, tanks, and rockets with as many chips as possible, enabling better guidance, communication, and command and control. In terms of producing military power, the strategy was working better than anyone except Bill Perry had thought possible.

只有一个问题。佩里曾认为诺伊斯和他的其他硅谷邻居将继续在行业中处于领先地位。但在 1986 年,日本的芯片产量已经超过美国。到 1980 年代末,日本供应了全球 70% 的光刻设备。美国的份额——在杰·拉斯罗普在美国军事实验室发明的一个行业——已经下降到 21%。光刻是“只是我们不能失去的东西,否则我们会发现自己完全依赖海外制造商来制造我们最敏感的东西,”一位国防部官员告诉《纽约时报》。但如果 1980 年代中期的趋势继续下去,日本将主导 DRAM 行业,并导致美国主要生产商倒闭。即使在阿拉伯禁运的最深处,美国可能会发现自己对外国芯片和半导体制造设备的依赖程度甚至超过了对石油的依赖。突然间,日本对其芯片行业的补贴被广泛指责为损害了英特尔和 GCA 等美国公司,这似乎是一个国家安全问题。

There was only one problem. Perry had assumed that Noyce and his other Silicon Valley neighbors would remain on top of the industry. But in 1986, Japan had overtaken America in the number of chips produced. By the end of the 1980s, Japan was supplying 70 percent of the world’s lithography equipment. America’s share—in an industry invented by Jay Lathrop in a U.S. military lab—had fallen to 21 percent. Lithography is “simply something we can’t lose, or we will find ourselves completely dependent on overseas manufacturers to make our most sensitive stuff,” one Defense Department official told the New York Times. But if the trends of the mid-1980s continued, Japan would dominate the DRAM industry and drive major U.S. producers out of business. The U.S. might find itself even more reliant on foreign chips and semiconductor manufacturing equipment than it was on oil, even at the depths of the Arab embargo. Suddenly Japan’s subsidies for its chip industry, widely blamed for undermining American firms like Intel and GCA, seemed like a national security issue.

国防部招募了杰克·基尔比、鲍勃·诺伊斯和其他行业名人来准备一份关于如何振兴美国半导体行业的报告。诺伊斯和基尔比在华盛顿郊区的头脑风暴会议上花费了数小时,与国防工业专家和五角大楼官员一起工作。鉴于德州仪器 (TI) 作为武器系统电子设备的主要供应商,Kilby 长期以来一直与国防部密切合作。IBM 和贝尔实验室也与华盛顿有着深厚的联系。但英特尔的领导人此前曾将自己描绘成正如一位国防官员所说,“不需要任何人帮助的硅谷牛仔”。诺伊斯愿意花时间在国防部的事实表明半导体行业面临的威胁——以及对美国军方的影响可能有多可怕。

The Defense Department recruited Jack Kilby, Bob Noyce, and other industry luminaries to prepare a report on how to revitalize America’s semiconductor industry. Noyce and Kilby spent hours at brainstorming sessions in the Washington suburbs, working with defense industrial experts and Pentagon officials. Kilby had long worked closely with the Defense Department, given Texas Instruments’ role as a major supplier of electronics for weapons systems. IBM and Bell Labs also had deep connections with Washington. But Intel’s leaders had previously portrayed themselves as “Silicon Valley cowboys who didn’t need anybody’s help,” as one defense official put it. The fact that Noyce was willing to spend time at the Defense Department was a sign of how serious a threat the semiconductor industry faced—and how dire the impact on the U.S. military could be.

美国军方比以往任何时候都更加依赖电子产品,因此也更加依赖芯片。报告发现,到 1980 年代,大约 17% 的军费开支用于电子产品,而二战结束时这一比例为 6%。从卫星到预警雷达再到自导导弹,一切都依赖于先进的芯片。五角大楼的工作组总结了四个要点,强调了关键结论:

The U.S. military was more dependent on electronics—and thus on chips—than ever before. By the 1980s, the report found, around 17 percent of military spending went toward electronics, compared to 6 percent at the end of World War II. Everything from satellites to early warning radars to self-guided missiles depended on advanced chips. The Pentagon’s task force summarized the ramifications in four bullet points, underlining the key conclusions:

  • 美国军队严重依赖技术优势 取胜
  • 电子技术是可以高度利用的技术。
  • 半导体是电子行业领先地位的关键。
  • 美国国防很快将依赖国外资源获取最先进的半导体技术。

当然,日本正式成为冷战盟友——至少目前如此。当美国在二战后不久占领日本时,它制定了日本宪法,使军国主义成为不可能。但在两国于1951年签署共同防御条约后,美国开始谨慎地鼓励日本重新武装,寻求对苏联的军事支持。东京同意,但将其军费开支限制在日本国内生产总值的 1% 左右。这是为了安抚日本的邻国,他们发自内心地记得日本的战时扩张主义。然而,由于日本并没有在军火上花费大量资金,所以它有更多的资金可以投资于其他地方。美国在国防上的花费是其经济规模的五到十倍。日本专注于发展经济,而美国则肩负起保卫经济的重任。

Of course, Japan was officially a Cold War ally—at least for now. When the U.S. had occupied Japan in the years immediately after World War II, it had written Japan’s constitution to make militarism impossible. But after the two countries had signed a mutual defense pact in 1951, the U.S. began cautiously to encourage Japanese rearmament, seeking military support against the Soviet Union. Tokyo agreed, but it capped its military spending around 1 percent of Japan’s GDP. This was intended to reassure Japan’s neighbors, who viscerally remembered the country’s wartime expansionism. However, because Japan didn’t spend heavily on arms, it had more funds to invest elsewhere. The U.S. spent five to ten times more on defense relative to the size of its economy. Japan focused on growing its economy, while America shouldered the burden of defending it.

结果比任何人预期的都要壮观。曾经被嘲笑为晶体管推销员的国家,日本现在是世界第二大经济体。这是对美国工业的挑战在对美国军事力量至关重要的领域占据主导地位。华盛顿长期以来一直敦促东京在日本扩大对外贸易的同时让美国遏制共产党,但这种分工似乎对美国不再非常有利。日本经济以前所未有的速度增长,而东京在高科技制造业方面的成功现在正威胁着美国的军事优势。日本的突进让所有人都措手不及。“你不希望半导体发生与电视行业、相机行业相同的事情,”Sporck 告诉五角大楼。“没有半导体,你就无处可去。”

The results were more spectacular than anyone had expected. Once derided as a country of transistor salesmen, Japan was now the world’s second-largest economy. It was challenging American industrial dominance in areas that were crucial to U.S. military power. Washington had long urged Tokyo to let the United States contain the Communists while Japan expanded its foreign trade, but this division of labor no longer seemed very favorable to the United States. Japan’s economy had grown at unprecedented speed, while Tokyo’s success in high-tech manufacturing was now threatening America’s military edge. Japan’s advance had caught everyone by surprise. “You don’t want the same thing to happen to semiconductors as happened to the TV industry, to the camera industry,” Sporck told the Pentagon. “Without semiconductors you’re in nowheresville.”

第十九章 死亡螺旋

CHAPTER 19 Death Spiral

我们正处于死亡螺旋中,”鲍勃·诺伊斯在 1986 年告诉记者。“你能说出美国没有落后的领域吗?” 在更悲观的时刻,诺伊斯想知道硅谷是否会像底特律一样结束,它的旗舰产业在外国竞争的影响下萎缩。硅谷与政府有着精神分裂的关系,同时要求不要管它,并要求它提供帮助。诺伊斯举例说明了这种矛盾。他早年在仙童度过,避开五角大楼的官僚主义,同时从冷战时期的太空竞赛中受益。现在他认为政府需要帮助半导体行业,但他仍然担心华盛顿会阻碍创新。与阿波罗计划时代不同,到 1980 年代超过 90% 的半导体是由公司和消费者购买的,而不是军方购买的。五角大楼很难塑造这个行业,因为国防部不再是硅谷最重要的客户。

“We’re in a death spiral,” Bob Noyce told a reporter in 1986. “Can you name a field in which the U.S. is not falling behind?” In his more pessimistic moments, Noyce wondered whether Silicon Valley would end up like Detroit, its flagship industry withering under the impact of foreign competition. Silicon Valley had a schizophrenic relationship with the government, simultaneously demanding to be left alone and requesting that it help. Noyce exemplified the contradiction. He’d spent his earliest days at Fairchild avoiding Pentagon bureaucracy while benefitting from the Cold War−era space race. Now he thought the government needed to help the semiconductor industry, but he still feared that Washington would impede innovation. Unlike in the days of the Apollo program, by the 1980s over 90 percent of semiconductors were bought by companies and consumers, not the military. It was hard for the Pentagon to shape the industry because the Defense Department was no longer Silicon Valley’s most important customer.

此外,在华盛顿,对于硅谷是否值得政府帮助几乎没有达成一致意见。毕竟,从汽车厂到钢铁厂,许多行业都在遭受日本的竞争。芯片行业和国防部认为半导体是“战略性的”。但许多经济学家认为,“战略”的含义并没有很好的定义。是半导体吗比喷气发动机更“战略”?还是工业机器人?“薯片,电脑芯片,有什么区别?” 一位里根政府的经济学家被广泛引用的话说。“他们都是筹码。一百元一元,一元一百元,还是一百元。” 这位经济学家否认曾将土豆与硅进行比较。但这是一个合理的观点。如果日本公司能够以更低的价格生产 DRAM 芯片,也许美国最好还是购买它们并将节省的成本收入囊中。如果是这样,那么美国的计算机会因此变得更便宜——并且计算机行业可能会发展得更快。

Moreover, in Washington there was little agreement on whether Silicon Valley merited government help. After all, many industries were suffering from Japanese competition, from car factories to steel mills. The chip industry and the Defense Department argued that semiconductors were “strategic.” But many economists argued that there was no good definition of what “strategic” meant. Were semiconductors more “strategic” than jet engines? Or industrial robots? “Potato chips, computer chips, what’s the difference?” one Reagan Administration economist was widely quoted as saying. “They’re all chips. A hundred dollars of one or a hundred dollars of the other is still a hundred.” The economist in question denies having ever compared potatoes to silicon. But the point was a reasonable one. If Japanese firms could produce DRAM chips at a lower price, perhaps the U.S. was better off buying them and pocketing the cost savings. If so, American computers would be cheaper as a result—and the computer industry might advance more quickly.

支持半导体的问题是由华盛顿游说决定的。硅谷和自由市场经济学家一致同意的一个问题是税收。鲍勃·诺伊斯向国会作证,支持将资本利得税从 49% 降至 28%,并主张放松金融监管,让养老基金投资于风险投资公司。在这些变化之后,大量资金涌入帕洛阿尔托沙丘路的风险投资公司。接下来,在英特尔的安迪·格鲁夫等硅谷高管向国会作证称日本公司的合法复制正在破坏美国的市场地位之后,国会通过《半导体芯片保护法》加强了知识产权保护。

The question of support for semiconductors was decided by lobbying in Washington. One issue on which Silicon Valley and free market economists agreed was taxes. Bob Noyce testified to Congress in favor of cutting the capital gains tax from 49 percent to 28 percent and advocated loosening financial regulation to let pension funds invest in venture capital firms. After these changes, a flood of money rushed into the venture capital firms on Palo Alto’s Sand Hill Road. Next, Congress tightened intellectual property protections via the Semiconductor Chip Protection Act, after Silicon Valley executives like Intel’s Andy Grove testified to Congress that legal copying by Japanese firms was undermining America’s market position.

然而,随着日本 DRAM 市场份额的增长,减税和版权变更似乎不足。五角大楼不愿将其国防工业基础押在版权法的未来影响上。硅谷的 CEO 们游说寻求更多帮助。诺伊斯估计,1980 年代他有一半的时间是在华盛顿度过的。杰里桑德斯抨击了日本一直追求的“补贴和培育、瞄准和保护市场”。“日本的补贴已达数十亿美元,”桑德斯宣称。即使在美国和日本达成取消半导体贸易关税的协议后,硅谷仍在努力向日本出售更多芯片。贸易谈判代表将与日本谈判比作剥洋葱。一位美国贸易谈判代表报告说:“整个事情是一种相当禅宗的体验,”讨论以诸如“什么是洋葱”之类的哲学问题。我们DRAM 对日本的销售几乎没有变化。

As Japan’s DRAM market share grew, however, tax cuts and copyright changes seemed insufficient. The Pentagon was unwilling to bet its defense industrial base on the future impact of copyright law. Silicon Valley CEOs lobbied for even more help. Noyce estimated that he spent half his time in the 1980s in Washington. Jerry Sanders attacked the “subsidies and nurturing, targeting and protection of markets” that Japan had pursued. “The Japanese subsidies have been in the billions,” Sanders declared. Even after the U.S. and Japan reached an agreement to eliminate tariffs on semiconductor trade, Silicon Valley struggled to sell Japan more chips. Trade negotiators compared negotiating with the Japanese to peeling an onion. “The whole thing is a rather zen experience,” one U.S. trade negotiator reported, with discussions ending with philosophical questions like “what is an onion, anyway.” U.S. DRAM sales into Japan barely budged.

在五角大楼的推动和工业界的游说下,里根政府最终决定采取行动。甚至像里根的国务卿乔治·舒尔茨这样的前自由贸易者也得出结论,只有在美国威胁要征收关税的情况下,日本才会开放市场。美国芯片行业对日本公司在美国市场“倾销”廉价芯片提出了一系列正式投诉。日本公司的销售低于生产成本的说法很难证明。美国公司引用日本竞争对手的低成本资本;日本回应说,日本经济的利率较低。双方都有道理。

Prodded by the Pentagon and lobbied by industry, the Reagan administration eventually decided to act. Even former free traders like Reagan’s secretary of state George Shultz concluded that Japan would only open its market if the U.S. threatened tariffs. America’s chip industry lodged a series of formal complaints against Japanese firms for “dumping” cheap chips in the U.S. market. The claim that Japanese firms were selling below production cost was hard to prove. U.S. firms cited Japanese competitors’ low cost of capital; Japan responded by saying that interest rates were lower across Japan’s economy. Both sides had a point.

1986 年,关税威胁迫在眉睫,华盛顿和东京达成协议。日本政府同意对其 DRAM 芯片的出口设置配额,限制销往美国的数量。通过减少供应,该协议推高了日本以外各地 DRAM 芯片的价格,损害了美国计算机生产商的利益。日本芯片的最大买家之一。更高的价格实际上使日本的生产商受益,他们继续主导着 DRAM 市场。大多数美国生产商已经在退出存储芯片市场的过程中。因此,尽管达成了贸易协议,但只有少数美国公司继续生产 DRAM 芯片。贸易限制重新分配了科技行业的利润,但它们无法拯救大多数美国存储芯片公司。

In 1986, with the threat of tariffs looming, Washington and Tokyo cut a deal. Japan’s government agreed to put quotas on its exports of DRAM chips, limiting the number that were sold to the U.S. By decreasing supply, the agreement drove up the price of DRAM chips everywhere outside of Japan, to the detriment of American computer producers, which were among the biggest buyers of Japan’s chips. Higher prices actually benefitted Japan’s producers, which continued to dominate the DRAM market. Most American producers were already in the process of exiting the memory chip market. So despite the trade deal, only a few U.S. firms continued to produce DRAM chips. The trade restrictions redistributed profits within the tech industry, but they couldn’t save most of America’s memory chip firms.

国会尝试了最后一种方式来提供帮助。硅谷的抱怨之一是日本政府帮助企业协调他们的研发工作并为此提供资金。美国高科技行业的许多人认为华盛顿应该复制这些策略。1987年,一群领先的芯片制造商和国防部创建了一个名为 Sematech 的财团,一半由工业界出资,一半由五角大楼出资。

Congress tried one final way to help. One of Silicon Valley’s complaints was that Japan’s government helped firms coordinate their R&D efforts and provided funds for this purpose. Many people in America’s high-tech industry thought Washington should replicate these tactics. In 1987, a group of leading chipmakers and the Defense Department created a consortium called Sematech, funded half by the industry and half by the Pentagon.

Sematech 的理念是行业需要更多的合作才能保持竞争力。芯片制造商需要更好制造设备,而生产这种设备的公司需要知道芯片制造商在寻找什么。设备公司的首席执行官抱怨说,“像 TI、摩托罗拉和 IBM 等公司……就是不愿公开他们的技术。” 如果不了解这些公司正在开发什么技术,就不可能向他们出售产品。与此同时,芯片制造商抱怨他们所依赖的机器的可靠性。在 1980 年代后期,由于维护和维修,英特尔的设备只有 30% 的时间运行,一名员工估计。

Sematech was based on the idea that the industry needed more collaboration to stay competitive. Chipmakers needed better manufacturing equipment, while the firms that produced this equipment needed to know what chipmakers were looking for. CEOs of equipment firms complained that “companies like TI, Motorola, and IBM… just would not open up about their technology.” Without an understanding of what technology these companies were working on, it was impossible to sell to them. Chipmakers, meanwhile, grumbled about the reliability of the machines they depended on. In the late 1980s, Intel’s equipment was running only 30 percent of the time due to maintenance and repairs, one employee estimated.

Bob Noyce 自愿领导 Sematech。事实上,他已经从英特尔退休,十年前将大权交给了戈登摩尔和安迪格罗夫。作为集成电路的共同发明者和美国最成功的两家初创公司的创始人,他拥有业内最好的技术和商业资格。没有人能比得上他的魅力或他在硅谷的人脉。如果说有谁能重振芯片行业,那一定是最有发言权的那个人。

Bob Noyce volunteered to lead Sematech. He was already de facto retired from Intel, having turned over the reins to Gordon Moore and Andy Grove a decade earlier. As the co-inventor of the integrated circuit and founder of two of America’s most successful startups, he had the best technical and business credentials in the industry. No one could match his charisma or his connections in Silicon Valley. If anyone could resuscitate the chip industry, it was the person with the strongest claim to have created it.

在诺伊斯的领导下,Sematech 是一个奇怪的混合体,既不是公司,也不是大学,也不是研究实验室。没有人确切知道它应该做什么。诺伊斯一开始试图帮助像 GCA 这样的制造设备公司,其中许多公司拥有强大的技术,但难以创造持久的业务或有效的制造流程。Sematech 组织了关于可靠性和良好管理技能的研讨会,提供了一种迷你 MBA 课程。它还开始在设备公司和芯片制造商之间进行协调,以调整他们的生产计划。如果光刻或沉积设备还没有准备好,那么芯片制造商准备新一代芯片制造技术就毫无意义。除非芯片制造商准备好使用它,否则设备公司不想推出新机器。Sematech 帮助他们就生产计划达成一致。这不完全是自由市场,但日本最大的公司在这种协调方面表现出色。不管怎样,硅谷还有什么选择?

Under Noyce’s leadership, Sematech was a strange hybrid, neither a company nor a university nor a research lab. No one knew exactly what it was supposed to do. Noyce started by trying to help manufacturing equipment companies like GCA, many of which had strong technology but struggled to create durable businesses or effective manufacturing processes. Sematech organized seminars on reliability and good management skills, offering a sort of mini-MBA. It also began coordinating between equipment companies and chipmakers to align their production schedules. There was no point in a chipmaker preparing a new generation of chipmaking technology if the lithography or deposition equipment wasn’t ready. Equipment firms didn’t want to launch a new piece of machinery unless chipmakers were prepared to use it. Sematech helped them agree on production schedules. This wasn’t exactly the free market, but Japan’s biggest firms had excelled with this type of coordination. Anyway, what other choice did Silicon Valley have?

然而,诺伊斯的重点是拯救美国的光刻行业。Sematech 51% 的资金流向了美国的光刻公司。诺伊斯简单地解释了这个逻辑:光刻机赚了一半的钱,因为它是芯片行业面临的“一半问题”。没有光刻工具就不可能制造半导体,但仅存的美国主要生产商正在努力生存。美国可能很快就会依赖外国设备。诺伊斯在 1989 年向国会作证时宣称,“Sematech 在很大程度上可能会受到评判,看它在拯救美国光学步进制造商方面的成功程度。”

Noyce’s focus, however, was saving America’s lithography industry. Fifty-one percent of Sematech funding went to American lithography firms. Noyce explained the logic simply: lithography got half the money because it was “half the problem” facing the chip industry. It was impossible to make semiconductors without lithography tools, but the only remaining major U.S. producers were struggling to survive. America might soon be reliant on foreign equipment. Testifying to Congress in 1989, Noyce declared that “Sematech may likely be judged, in large part, as to how successful it is in saving America’s optical stepper makers.”

这正是生病的马萨诸塞州光刻工具制造商 GCA 的员工希望听到的。在公司发明了晶圆步进机之后,五年的管理不善和运气不佳使 GCA 成为一个小玩家,远远落后于日本的尼康和佳能以及荷兰的 ASML。但是当 GCA 的总裁 Peter Simone 打电话给 Noyce 讨论 Sematech 是否可以帮助 GCA 时,Noyce 直截了当地告诉他:“你完成了。”

This was exactly what employees at GCA, the ailing Massachusetts manufacturer of lithography tools, were hoping to hear. After the company had invented the wafer stepper, a half decade of mismanagement and bad luck had left GCA a small player, far behind Japan’s Nikon and Canon and the Netherlands’ ASML. But when Peter Simone, GCA’s president, called Noyce to discuss whether Sematech could help GCA, Noyce told him flatly: “You’re done.”

芯片行业很少有人能看到 GCA 是如何复苏的。诺伊斯创立的英特尔严重依赖 GCA 在日本的主要竞争对手尼康。“你为什么不来一天,”西蒙尼提议,希望能说服诺伊斯,GCA 仍然可以生产尖端机械。诺伊斯同意了,当他到达马萨诸塞州时他当天决定购买价值 1300 万美元的 GCA 最新设备,作为与美国芯片制造商共享美国制造的半导体设备并鼓励他们购买更多国产工具的计划的一部分。

Few people in the chip industry could see how GCA could recover. Intel, which Noyce had founded, relied heavily on Nikon, GCA’s primary Japanese competitor. “Why don’t you come for one day,” Simone proposed, hoping to convince Noyce that GCA could still produce cutting-edge machinery. Noyce agreed, and when he arrived in Massachusetts he decided that day to buy $13 million worth of GCA’s newest equipment, as part of a program to share American-built semiconductor equipment with U.S. chipmakers and encourage them to buy more domestically produced tools.

Sematech 对 GCA 下了巨大的赌注,为该公司提供了生产处于行业前沿能力的深紫外光刻设备的合同。GCA 的交付远远超出预期,不辜负其早先在技术卓越方面的声誉。很快,独立的行业分析师将 GCA 的最新步进电机描述为“世界上最好的”。该公司甚至赢得了客户服务奖,摆脱了该部门平庸的声誉。GCA的机器使用的软件要好得多超过该公司的日本竞争对手。“他们领先于时代,”德州仪器 (TI) 一位测试 GCA 最新机器的光刻专家回忆道。

Sematech bet hugely on GCA, giving the company contracts to produce deep-ultraviolet lithography equipment that was at the cutting edge of the industry’s capabilities. GCA delivered far beyond expectations, living up to its earlier reputation for technological brilliance. Soon independent industry analysts were describing GCA’s newest steppers as “the best in the world.” The company even won a customer service award, casting off its reputation for being mediocre in that department. The software that GCA’s machines used was far better than the company’s Japanese rivals. “They were ahead of their time,” recalled one lithography expert at Texas Instruments who tested GCA’s newest machines.

但 GCA 仍然没有可行的商业模式。“领先于你的时间”对科学家来说是好事,但对寻求销售的制造公司来说却不一定。客户已经对来自尼康、佳能和 ASML 等竞争对手的设备感到满意,并且不想在未来不确定的公司提供的新的和不熟悉的工具上冒险。如果 GCA 破产,客户可能很难获得备件。除非能够说服大客户与 GCA 签署一份重大合同,否则该公司将走向崩溃。尽管 Sematech 提供了 7000 万美元的支持,但它在 1988 年至 1992 年间损失了 3000 万美元。即使是诺伊斯也无法说服他创立的公司英特尔从尼康转而效忠。

But GCA still didn’t have a viable business model. Being “ahead of your time” is good for scientists but not necessarily for manufacturing firms seeking sales. Customers had already gotten comfortable with equipment from competitors like Nikon, Canon, and ASML, and didn’t want to take a risk on new and unfamiliar tools from a company whose future was uncertain. If GCA went bankrupt, customers might struggle to get spare parts. Unless a big customer could be convinced to sign a major contract with GCA, the company would spiral toward collapse. It lost $30 million between 1988 and 1992, despite $70 million in support from Sematech. Even Noyce could never convince Intel, the company he’d founded, to switch its allegiance from Nikon.

1990 年,GCA 在 Sematech 的最大支持者诺伊斯在早上游泳后死于心脏病发作。他创建了仙童和英特尔,发明了集成电路,并将支撑所有现代计算的 DRAM 芯片和微处理器商业化。然而,光刻技术证明对诺伊斯的魔法免疫。到 1993 年,GCA 的所有者,一家名为 General Signal 的公司,宣布将出售或关闭 GCA。随着时间快到这个自我设定的最后期限,找不到买家。已经为 GCA 提供了数百万美元资金的 Sematech 决定停止运营。GCA 最后一次向政府寻求帮助,国家安全高级官员正在考虑美国的外交政策是否需要拯救 GCA。他们得出结论无能为力。该公司关门大吉,卖掉了设备,加入了一长串被日本竞争击败的公司行列。

In 1990, Noyce, GCA’s greatest supporter at Sematech, died of a heart attack after his morning swim. He’d built Fairchild and Intel, invented the integrated circuit, and commercialized the DRAM chips and microprocessors that undergird all modern computing. Lithography, however, proved immune to Noyce’s magic. By 1993, GCA’s owner, a company called General Signal, announced it would sell GCA or close it. As the clock ticked toward this self-imposed deadline, no buyer could be found. Sematech, which had already provided millions in funding for GCA, decided to pull the plug. GCA appealed one final time to the government for help, with top national security officials considering whether U.S. foreign policy required saving GCA. They concluded nothing could be done. The company shut its doors and sold off its equipment, joining a long list of firms vanquished by Japanese competition.

第 20 章 可以说不的日本

CHAPTER 20 The Japan That Can Say No

通过向美国人销售电子产品赚取数百万美元之后,索尼的Akio Morita 开始在他的美国朋友身上发现“某种傲慢”。当他在 1950 年代首次获得晶体管技术许可时,美国是世界技术领导者。从那时起,美国面临着一场又一场的危机。越南的灾难性战争、种族紧张局势、城市动荡、水门事件的耻辱、十年的滞胀、巨大的贸易逆差,以及现在的工业萎靡不振。在每一次新的冲击之后,美国的吸引力就减弱了。

After decades of making millions by selling Americans electronics, Sony’s Akio Morita began to detect “a certain arrogance” in his American friends. When he first licensed transistor technology in the 1950s, the U.S. was the world’s tech leader. Since then, America had faced crisis after crisis. The disastrous war in Vietnam, racial tension, urban unrest, the humiliation of Watergate, a decade of stagflation, a gaping trade deficit, and now industrial malaise. After each new shock, America’s allure dimmed.

1953 年他第一次出国旅行时,森田将美国视为一个“似乎拥有一切”的国家。他得到了冰淇淋,上面放着一把小纸伞。“这是从你的国家来的,”服务员告诉他,这是一种耻辱,提醒他日本落后了多么远。然而,三十年后,一切都变了。森田在 1950 年代首次访问时,纽约似乎“魅力四射”。现在它肮脏,犯罪猖獗,破产了。

On his first trip abroad in 1953, Morita had seen America as a country “that seemed to have everything.” He was served ice cream with a tiny paper umbrella on the top. “This is from your country,” the waiter told him, a humiliating reminder of how far behind Japan was. Three decades later, however, everything had changed. New York had seemed “glamorous” on Morita’s first visit in the 1950s. Now it was dirty, crime-ridden, and bankrupt.

与此同时,索尼已经成为一个全球品牌。森田重新定义了日本在海外的形象。该国不再被视为冰淇淋圣代的纸伞生产国。现在它制造了世界上最高科技的产品。森田的家族拥有索尼的主要股份,因此变得富有。他在华尔街和华盛顿拥有强大的朋友网络。他培养了纽约晚宴的艺术作为当其他日本人走近传统的茶道时,他一丝不苟。每当森田在纽约时,他都会在大都会艺术博物馆对面的 82 号和 5 号公寓接待该市的富豪和名人。森田的妻子芳子甚至写了一本书,向不熟悉的日本读者解释美国的晚宴习俗,名为《我对家庭娱乐的思考》。(不鼓励穿和服;“只要每个人都穿同样的衣服,和谐就会增强。”)

Sony, meanwhile, had become a global brand. Morita redefined Japan’s image abroad. The country was no longer seen as a producer of paper umbrellas for ice cream sundaes. Now it built the world’s most high-tech goods. Morita, whose family owned a major stake in Sony, had gotten rich. He had a powerful network of friends on Wall Street and in Washington. He cultivated the art of the New York dinner party as meticulously as other Japanese approached a traditional tea ceremony. Whenever Morita was in New York, he hosted the city’s rich and famous at his apartment on 82nd and Fifth, just across from the Metropolitan Museum of Art. Morita’s wife Yoshiko even wrote a book explaining American dinner party customs to unfamiliar Japanese readers, titled My Thoughts on Home Entertaining. (Kimonos were discouraged; “whenever everyone wears the same kind of outfit, harmony is enhanced.”)

森田夫妇喜欢娱乐,但他们的晚宴也有专业目的。随着美国和日本之间的商业紧张局势加剧,森田担任非正式大使,向美国的权力掮客解释日本。大卫洛克菲勒是一位私人朋友。每当这位前国务卿访问日本时,森田都会与亨利·基辛格共进晚餐。当私募股权巨头皮特·彼得森将森田带到深受 CEO 欢迎的高尔夫俱乐部奥古斯塔时,他震惊地发现“Akio 已经认识了他们所有人”。不仅如此——森田在奥古斯塔与他的每个熟人都安排了晚餐。“他一定有过他住在这里的时候一天吃十顿饭,”彼得森回忆道。

The Moritas enjoyed entertaining, but their dinner parties served a professional purpose, too. As commercial tension between the U.S. and Japan increased, Morita served as informal ambassador, explaining Japan to American powerbrokers. David Rockefeller was a personal friend. Morita dined with Henry Kissinger whenever the former secretary of state visited Japan. When private equity titan Pete Peterson took Morita to Augusta National, a golf club popular with CEOs, he was shocked to discover that “Akio had met them all.” Not only that—Morita arranged a dinner with each of his acquaintances while at Augusta. “He must have had about ten meals a day while he was staying here,” Peterson recounted.

森田起初觉得他的美国朋友所代表的权力和财富很诱人。然而,随着美国从危机到危机,亨利·基辛格和皮特·彼得森等人的光环开始减弱。他们国家的系统不起作用——但是日本的曾经。到 1980 年代,森田意识到美国经济和社会存在严重问题。长期以来,美国一直将自己视为日本的老师,但森田认为美国在与不断增长的贸易逆差和高科技产业危机作斗争时可以吸取教训。“美国一直在忙于培养律师,”森田说,而日本“一直在忙于培养工程师”。此外,美国高管过于关注“今年的利润”,而日本的管理层则“长期”。美国的劳资关系是等级森严的和“旧式”的,对车间员工没有足够的培训或动力。森田认为,美国人应该停止抱怨日本的成功。是时候告诉他的美国朋友了:日本的制度运作得更好。

Morita at first found the power and wealth represented by his American friends seductive. As America lurched from crisis to crisis, however, the aura around men like Henry Kissinger and Pete Peterson began to wane. Their country’s system wasn’t working—but Japan’s was. By the 1980s, Morita perceived deep problems in America’s economy and society. America had long seen itself as Japan’s teacher, but Morita thought America had lessons to learn as it struggled with a growing trade deficit and the crisis in its high-tech industries. “The United States has been busy creating lawyers,” Morita lectured, while Japan has “been busier creating engineers.” Moreover, American executives were too focused on “this year’s profit,” in contrast to Japanese management, which was “long range.” American labor relations were hierarchical and “old style,” without enough training or motivation for shop-floor employees. Americans should stop complaining about Japan’s success, Morita believed. It was time to tell his American friends: Japan’s system simply worked better.

1989 年,森田在题为《可以说不的日本为什么日本将在平等中居首位》的论文集中阐述了他的观点。这本书是与备受争议的极右翼政治家石原慎太郎合着的。石原只是一名大学生,就因出版了一部名为《太阳的季节,被授予日本最负盛名的新作家文学奖。作为执政的自民党成员,他利用这种因对外国人的贬损而增强的名声,成为议会席位。在议会中,石原鼓动日本在国际上坚持自己的立场,并修改二战后美国占领当局制定的国家宪法,让东京建立一支强大的军队。

In 1989, Morita set out his views in a collection of essays titled The Japan That Can Say No: Why Japan Will Be First Among Equals. The book was coauthored with Shintaro Ishihara, a controversial far-right politician. While just a university student, Ishihara had risen to fame by publishing a sexually charged novel titled Season of the Sun, which was awarded Japan’s most prestigious literary prize for new writers. He parlayed this fame, enhanced by derogatory diatribes against foreigners, into a parliamentary seat as a member of the ruling Liberal Democratic Party. In parliament, Ishihara agitated for Japan to assert itself internationally and to change the country’s constitution, which had been dictated by U.S. occupation authorities after World War II, to let Tokyo build a powerful military.

很难想象森田在向美国讲述其内部危机时选择了更具挑衅性的合著者。这本书本身就是一系列散文,一些是森田写的,另一些是石原写的。森田的文章大多重复了他关于美国商业实践失败的论点,尽管诸如“美国,你最好放弃某些傲慢”之类的章节标题比森田通常在纽约晚宴上表达的语气更严厉。即使是一向和蔼可亲的森田也难以掩饰他的观点,即日本的技术实力已经为其在世界大国中赢得了一席之地。“在军事上,我们永远无法击败美国,”森田当时对一位美国同事说,“但在经济上,我们可以战胜美国,成为世界第一。”

It was hard to imagine a more provocative coauthor for Morita to have chosen as he lectured the United States about its internal crises. The book itself was a series of essays, some written by Morita and others by Ishihara. Morita’s essays mostly rehashed his arguments about the failings of American business practices, though chapter titles such as “America, You Had Better Give Up Certain Arrogance” had a harsher tone than Morita usually expressed at New York dinner parties. Even the always gracious Morita found it difficult to mask his view that Japan’s technological prowess had earned it a position among the world’s great powers. “Militarily we could never defeat the United States,” Morita told an American colleague at the time, “but economically we can overcome the United States and become number one in the world.”

石原毫不犹豫地说出了自己的想法。他的第一部小说是关于不受约束的性冲动的故事。他的政治生涯包含了日本民族主义最令人讨厌的本能。他在《可以说不的日本》中的文章呼吁日本宣布独立,摆脱霸道日本太久的霸道美国。“我们不要屈服于美国的咆哮!” 石原的一篇文章宣称。“遏制美国!” 宣布另一个。日本的极右翼一直对他们的国家在美国主导的世界中的次要地位感到不满。森田愿意合着一本书像石原这样的人震惊了许多美国人,表明威胁性的民族主义仍然潜伏在华盛顿培养的资产阶级中。自 1945 年以来,美国的战略一直是通过贸易和技术交流将日本与美国联系起来。Akio Morita 可以说是美国技术转让和市场开放的最大受益者。即使在质疑美国的领导作用,华盛顿也需要重新考虑其游戏计划。

Ishihara never hesitated to say exactly what he was thinking. His first novel was a story of unconstrained sexual urges. His political career embraced the most unsavory instincts of Japanese nationalism. His essays in The Japan That Can Say No called for Japan to declare independence from an overbearing America that had bossed Japan around for too long. “Let’s not give into America’s bluster!” one of Ishihara’s essays proclaimed. “Restrain America!” declared another. Japan’s far right had always been unhappy with their country’s secondary status in an America-led world. Morita’s willingness to coauthor a book with someone like Ishihara shocked many Americans, showing that a threatening nationalism still lurked within the capitalist class that Washington had cultivated. The U.S. strategy since 1945 had been to bind Japan to the U.S. via exchanges of trade and technology. Akio Morita was arguably the greatest beneficiary of America’s tech transfers and its market openness. If even he was questioning America’s leading role, Washington needed to rethink its game plan.

《可以说不的日本》真正让华盛顿感到恐惧的原因不仅在于它表达了一种零和的日本民族主义,而且石原已经找到了一种胁迫美国的方法。石原认为,日本不需要服从美国的要求,因为美国依赖日本的半导体。他指出,美国的军事实力需要日本的筹码。他写道:“无论是中程核武器还是洲际弹道导弹,确保武器准确性的无非是紧凑型高精度计算机。” “如果不使用日本半导体,就无法保证这种准确性。” 石原推测日本甚至可以向苏联提供先进的半导体,从而打破冷战中的军事平衡。

What made The Japan That Can Say No truly frightening to Washington was not only that it articulated a zero-sum Japanese nationalism, but that Ishihara had identified a way to coerce America. Japan didn’t need to submit to U.S. demands, Ishihara argued, because America relied on Japanese semiconductors. American military strength, he noted, required Japanese chips. “Whether it be mid-range nuclear weapons or inter-continental ballistic missiles, what ensures the accuracy of weapons is none other than compact, high-precision computers,” he wrote. “If Japanese semiconductors are not used, this accuracy cannot be assured.” Ishihara speculated that Japan could even provide advanced semiconductors to the USSR, tipping the military balance in the Cold War.

“用于计算机核心的 1 兆位半导体,在你小指甲三分之一大小的区域内承载数亿个电路,仅在日本制造,”石原说。“日本拥有这些 1 兆位半导体的近 100% 份额。“现在日本在这方面至少领先美国五年,而且差距正在扩大,”他继续说道。使用日本芯片的计算机“是军事实力的核心,因此也是日本实力的核心……从这个意义上说,日本已成为一个非常重要的国家。”

“The 1-megabit semiconductors which are used in the hearts of computers, which carry hundreds of millions of circuits in an area which is one-third the size of your little fingernail, are only made in Japan,” Ishihara noted. “Japan has nearly a 100 percent share of these 1-megabit semiconductors. “Now Japan is at least five years ahead of the U.S. in this area and the gap is widening,” he continued. Computers using Japan’s chips were “central to military strength and therefore central to Japanese power… in that sense, Japan has become a very important country.”

其他日本领导人似乎也采取了同样挑衅的民族主义观点。一位外交部高级官员被引述称,“美国人根本不想承认日本在与西方的经济竞赛中获胜。” 即将成为日本首相的宫泽喜一公开指出,切断日本电子产品出口将导致“美国经济出现问题”,并预测“亚洲经济区将超过北美区”。当中一位日本教授宣称,美国的工业和高科技部门崩溃,美国的未来是“一个首要的农业强国,一个丹麦的巨型版本。”

Other Japanese leaders appeared to take a similarly defiant nationalist view. One senior Foreign Ministry official was quoted as arguing that “Americans simply don’t want to recognize that Japan has won the economic race against the West.” Soon-to-be-prime-minister Kiichi Miyazawa publicly noted that cutting off Japanese electronics exports would cause “problems in the U.S. economy,” and predicted that “the Asian economic zone will outdo the North American zone.” Amid the collapse of its industries and its high-tech sector, America’s future, a Japanese professor declared, was that of “a premier agrarian power, a giant version of Denmark.”

在美国,可以说不的日本引发了愤怒。它由中央情报局以非官方形式翻译和传播。一位愤怒的国会议员将整本书——仍然只是非正式地以英文出版——进入国会记录以进行宣传。书店报告说华盛顿的顾客“疯了”试图找到盗版。森田不好意思地让官方英文翻译只与石原的论文一起出版,没有他的贡献。“我现在后悔与这个项目有联系,”森田告诉记者,“因为它造成了很多混乱。我觉得美国读者不明白我的观点与石原的观点是不同的。我的‘散文’表达我的观点,他的‘散文’表达他的观点。”

In the U.S., The Japan That Can Say No sparked fury. It was translated and circulated in unofficial form by the CIA. One irate congressman entered the entire book—still published in English only unofficially—into the Congressional Record to publicize it. Bookstores reported that customers in Washington were “going absolutely bananas” trying to find bootleg copies. Morita sheepishly had the official English translation published only with Ishihara’s essays, without his contributions. “I now regret my association with this project,” Morita told reporters, “because it has caused so much confusion. I don’t feel U.S. readers understand that my opinions are separate from Ishihara’s. My ‘essays’ express my opinions and his ‘essays’ express his opinions.”

然而,《可以说不的日本》之所以引起争议,不是因为它的观点,而是因为事实。美国在存储芯片方面已经明显落后。如果这种趋势持续下去,地缘政治的转变将不可避免地随之而来。像石原这样的极右翼挑衅者并没有意识到这一点。美国领导人预见到了类似的趋势。石原和森田发表《可以说不的日本》的同一年,前国防部长哈罗德·布朗发表了一篇文章,得出了大致相同的结论。“高科技就是外交政策,”布朗为这篇文章命名。如果美国的高科技地位在恶化,它的外交政策地位也将处于危险之中。

Yet The Japan That Can Say No was controversial not because of its opinions, but because of the facts. The U.S. had fallen decisively behind in memory chips. If this trend persisted, geopolitical shifts would inevitably follow. It didn’t take a far-right provocateur like Ishihara to recognize this; American leaders foresaw similar trends. The same year that Ishihara and Morita published The Japan That Can Say No, former defense secretary Harold Brown published an article that drew much the same conclusions. “High Tech Is Foreign Policy,” Brown titled the article. If America’s high-tech position was deteriorating, its foreign policy position was at risk, too.

对于五角大楼领导人布朗来说,这是一个令人尴尬的承认,他在 1977 年聘请了比尔佩里,并授权他将半导体和计算能力置于军队最重要的新武器系统的核心。布朗和佩里成功说服军方接受微处理器,但他们没有预料到硅谷会失去领先地位。他们的战略在新武器系统方面得到了回报,但其中许多现在依赖于日本。

This was an embarrassing admission for Brown, the Pentagon leader who’d hired Bill Perry in 1977 and empowered him to put semiconductors and computing power at the core of the military’s most important new weapons systems. Brown and Perry succeeded in convincing the military to embrace microprocessors, but they hadn’t anticipated Silicon Valley losing its lead. Their strategy paid off in terms of new weapons systems, but many of these now depended on Japan.

“日本在存储芯片方面处于领先地位,这是消费电子产品的核心,”布朗承认。“日本人正在迅速追赶在逻辑芯片和专用集成电路中。” 日本还在制造芯片所需的某些类型的工具方面处于领先地位,例如光刻设备。布朗可以预见的最好结果是美国将保护日本的未来,但将使用由日本技术驱动的武器来保护日本。美国将日本变成晶体管推销员的战略似乎大错特错。

“Japan leads in memory chips, which are at the heart of consumer electronics,” Brown admitted. “The Japanese are rapidly catching up in logic chips and application-specific integrated circuits.” Japan also led in certain types of tools, like lithography equipment, needed to build chips. The best result Brown could foresee was a future in which the U.S. would protect Japan, but would do so with weapons powered by Japanese tech. America’s strategy to turn Japan into a transistor salesman seemed to have gone horribly wrong.

技术一流的日本,会满足于二流的军事地位吗?如果日本在 DRAM 芯片方面的成功可以作为参考,那么它几乎在所有重要的行业都将超过美国。为什么它不寻求军事统治呢?如果是这样,美国会怎么做?1987 年,中央情报局委托一个分析师团队预测亚洲的未来。他们将日本在半导体领域的主导地位视为新兴的“日本和平”——由日本领导的东亚经济和政治集团——的证据。美国在亚洲的实力建立在技术优势、军事实力以及将日本、香港、韩国和东南亚国家联系在一起的贸易和投资联系之上。从位于香港九龙湾的第一家仙童组装厂开始,集成电路就一直是美国在亚洲地位不可或缺的特征。美国芯片制造商在台湾、韩国、新加坡等地建造了设施。这些领土不仅通过军事力量而且还通过经济一体化来抵御共产党的入侵,

Would Japan, a first-class technological power, be satisfied with second-class military status? If Japan’s success in DRAM chips was any guide, it was set to overtake the United States in almost every industry that mattered. Why wouldn’t it seek military dominance, too? If so, what would the U.S. do? In 1987, the CIA tasked a team of analysts with forecasting Asia’s future. They saw Japanese dominance of semiconductors as evidence of an emerging “Pax Niponica”—an East Asian economic and political bloc led by Japan. American power in Asia had been built on technological dominance, military might, and trade and investment links that knit together Japan, Hong Kong, South Korea, and the countries of Southeast Asia. From the first Fairchild assembly plant on Hong Kong’s Kowloon Bay, integrated circuits had been an integral feature of America’s position in Asia. U.S. chipmakers built facilities from Taiwan to South Korea to Singapore. These territories were defended from Communist incursions not only by military force but also by economic integration, as the electronics industry sucked the region’s peasants off farms—where rural poverty often inspired guerilla opposition—into good jobs assembling electronic devices for American consumption.

美国的供应链治国之道在抵御共产党方面表现出色,但到了 1980 年代,主要受益者似乎是日本。它的贸易和外国投资大幅增长。东京在亚洲经济和政治中的作用正在无情地扩大。如果日本能够如此迅速地确立对芯片行业的主导地位,那么又如何阻止它取代美国的地缘政治优势呢?

America’s supply chain statecraft had worked brilliantly in fending off Communists, but by the 1980s, the primary beneficiary looked to have been Japan. Its trade and foreign investment had grown massively. Tokyo’s role in Asia’s economics and politics was expanding inexorably. If Japan could so swiftly establish dominance over the chip industry, what would stop it from dethroning America’s geopolitical preeminence, too?

第四部分 美国复兴

PART IV AMERICA RESURGENT

第 21 章 薯片王

CHAPTER 21 The Potato Chip King

杰克·辛普洛特曾经说过,微米制造了“全世界最好的小部件”。这位爱达荷州的亿万富翁对他公司的主要产品 DRAM 芯片的实际工作原理知之甚少。芯片行业充满了博士,但 Simplot 还没有读完八年级。他的专长是土豆,众所周知,他开着白色的林肯城市汽车在博伊西附近行驶。“先生。Spud,”车牌宣称。然而,辛普洛对商业的理解却是硅谷最聪明的科学家所没有的。当美国的芯片行业努力适应日本的挑战时,像他这样的牛仔企业家在扭转 Bob Noyce 所说的“死亡螺旋”并实现意外转变方面发挥了重要作用。

Micron made “the best damn widgets in the whole world,” Jack Simplot used to say. The Idaho billionaire didn’t know much about the physics of how his company’s main product, DRAM chips, actually worked. The chip industry was full of PhDs, but Simplot hadn’t finished eighth grade. His expertise was potatoes, as everyone knew from the white Lincoln Town Car he drove around Boise. “Mr. Spud,” the license plate declared. Yet Simplot understood business in a way Silicon Valley’s smartest scientists didn’t. As America’s chip industry struggled to adjust to Japan’s challenge, cowboy entrepreneurs like him played a fundamental role in reversing what Bob Noyce had called a “death spiral” and executing a surprise turnaround.

硅谷的复苏是由斗志旺盛的初创公司和痛苦的企业转型推动的。美国超越日本的 DRAM 庞然大物不是通过复制它们,而是通过围绕它们进行创新。硅谷没有切断自己的贸易,而是将更多的生产转移到台湾和韩国,以重新获得竞争优势。与此同时,随着美国芯片业的复苏,五角大楼在微电子领域的赌注开始得到回报,因为它部署了其他国家无法匹敌的新武器系统。美国在 1990 年代和 2000 年代无与伦比的实力源于其在计算机芯片这一时代核心技术领域的重新主导地位。

Silicon Valley’s resurgence was driven by scrappy startups and by wrenching corporate transformations. The U.S. overtook Japan’s DRAM behemoths not by replicating them but by innovating around them. Rather than cutting itself off from trade, Silicon Valley offshored even more production to Taiwan and South Korea to regain its competitive advantage. Meanwhile, as America’s chip industry recovered, the Pentagon’s bet on microelectronics began to pay off as it fielded new weapons systems that no other country could match. America’s unrivaled power during the 1990s and 2000s stemmed from its resurgent dominance in computer chips, the core technology of the era.

在所有帮助重振美国芯片产业的人中,杰克·辛普洛是最不可能的候选人。他在土豆上发了第一笔财富,开创了使用机器对土豆进行分类、脱水和冷冻以用于炸薯条的先河。这不是硅谷式的创新,但它为他赢得了一份向麦当劳出售土豆的巨额合同。有一次,他提供了麦当劳用来做薯条的一半土豆。

Of all the people to help revive America’s chip industry, Jack Simplot was the least likely candidate. He’d made his first fortune in potatoes, pioneering the use of machines to sort potatoes, dehydrate them, and freeze them for use in french fries. This wasn’t Silicon Valley−style innovation, but it earned him a massive contract to sell spuds to McDonald’s. At one point he supplied half the potatoes that McDonald’s used to make fries.

Simplot 支持的 DRAM 公司美光起初似乎注定要失败。1978 年,双胞胎兄弟乔和沃德·帕金森在博伊西牙医诊所的地下室创立了美光科技,那是创办存储芯片公司的最糟糕的时期。日本公司正在加大生产高质量、低价位的内存芯片的力度。美光的第一份合同是为德克萨斯州一家名为 Mostek 的公司设计 64K DRAM 芯片,但与其他所有美国 DRAM 生产商一样,它被富士通抢先进入市场。很快,Mostek——美光芯片设计服务的唯一客户——破产了。在日本的激烈竞争中,AMD、美国国家半导体、英特尔和其他行业领导者也放弃了 DRAM 生产。面对数十亿美元的损失和破产,整个硅谷似乎都可能破产。美国最聪明的工程师将被丢在汉堡上。至少,这个国家还有很多薯条。

Micron, the DRAM firm that Simplot backed, at first seemed guaranteed to fail. When twin brothers Joe and Ward Parkinson founded Micron in the basement of a Boise dentist office in 1978, it was the worst possible time to start a memory chip company. Japanese firms were ramping up production of high-quality, low-priced memory chips. Micron’s first contract was to design a 64K DRAM chip for a Texas company called Mostek, but like every other American DRAM producer, it was beaten to the market by Fujitsu. Soon Mostek—the only customer for Micron’s chip design services—went bust. Amid an onslaught of Japanese competition, AMD, National Semiconductor, Intel, and other industry leaders abandoned DRAM production, too. Facing billion-dollar losses and bankruptcies, it seemed like all Silicon Valley might go bankrupt. America’s smartest engineers would be left flipping burgers. At least, the country still had plenty of french fries.

随着日本公司抢占市场份额,美国最大的芯片公司的首席执行官们在华盛顿花费了越来越多的时间,游说国会和五角大楼。在日本竞争加剧的那一刻,他们搁置了自由市场的信念,声称竞争是不公平的。硅谷愤怒地拒绝了薯片和电脑芯片没有区别的说法。他们坚持认为,他们的筹码值得政府帮助,因为它们具有战略意义,而土豆则不然。

As Japanese firms grabbed market share, CEOs of America’s biggest chip firms spent more and more time in Washington, lobbying Congress and the Pentagon. They set aside their free-market beliefs the moment Japanese competition mounted, claiming the competition was unfair. Silicon Valley angrily rejected the claim that there was no difference between potato chips and computer chips. Their chips merited government help, they insisted, because they were strategic in a way spuds weren’t.

杰克辛普洛特没有看到土豆有什么问题。硅谷应该得到特别帮助的论点在爱达荷州并没有走得太远,这个州几乎没有科技公司。美光不得不以艰难的方式筹集资金。美光联合创始人沃德·帕金森(Ward Parkinson)认识了一个博伊西商人艾伦·诺布尔(Allen Noble)穿着西装穿过诺布尔泥泞的马铃薯田,试图找到灌溉系统中出现故障的电子元件。帕金森兄弟利用这种联系从 Noble 和他的几个富有的博伊西朋友那里获得了 100,000 美元的种子资金。当美光失去为 Mostek 设计芯片的合同并决定制造自己的芯片时,帕金森一家需要更多资金。于是他们求助于州首富斯普德先生。

Jack Simplot didn’t see anything wrong with potatoes. The argument that Silicon Valley deserved special help didn’t go very far in Idaho, a state with few tech companies. Micron had had to raise funds the hard way. Micron cofounder Ward Parkinson had gotten to know a Boise businessman named Allen Noble when he waded through Noble’s muddy potato field in a business suit trying to find a malfunctioning electric component in an irrigation system. The Parkinson brothers parlayed this connection into $100,000 in seed funding from Noble and a couple of his wealthy Boise friends. When Micron lost its contract to design chips for Mostek and decided to make its own chips, the Parkinsons needed more capital. So they turned to Mr. Spud, the richest man in the state.

帕金森兄弟在博伊西市中心的皇家咖啡厅第一次见到辛普洛特,当他们向爱达荷州的马铃薯富豪投球时,他们大汗淋漓。晶体管和电容器对辛普洛来说意义不大,他与硅谷风险投资家的对立面是你所能得到的。他后来每周一早上 5 点 45 分在 Elmer's 主持即兴的美光董事会会议,一个当地油腻的勺子,供应成堆的酪乳煎饼,售价 6.99 美元。然而,在日本的猛攻中,所有硅谷的科技巨头都在逃离 DRAM 芯片,辛普洛本能地明白,沃德和乔·帕金森进入内存市场的时机恰到好处。像他这样的马铃薯农民清楚地看到,日本的竞争已经将 DRAM 芯片变成了商品市场。他经历了足够多的收获,知道购买商品业务的最佳时机是价格低迷且其他所有人都处于清算状态的时候。Simplot 决定以 100 万美元支持美光。他后来再注入数百万。

The Parkinson brothers first met Simplot at the Royal Café in downtown Boise, pouring sweat as they delivered their pitch to Idaho’s potato plutocrat. Transistors and capacitors didn’t mean much to Simplot, who was as close to the opposite of a Silicon Valley venture capitalist as you could get. He’d later preside over impromptu Micron board meetings each Monday at 5:45 a.m. at Elmer’s, a local greasy spoon that served stacks of buttermilk pancakes for $6.99. However, as all of Silicon Valley’s tech titans were fleeing DRAM chips amid the Japanese onslaught, Simplot instinctively understood that Ward and Joe Parkinson were entering the memory market at exactly the right time. A potato farmer like him saw clearly that Japanese competition had turned DRAM chips into a commodity market. He’d been through enough harvests to know that the best time to buy a commodity business was when prices were depressed and everyone else was in liquidation. Simplot decided to back Micron with $1 million. He’d later pour in millions more.

美国的科技巨头认为爱达荷州的乡巴佬一点头绪都没有。“我不想说内存芯片已经结束了,”LJ Sevin 说,他是德州仪器 (Texas Instruments) 的前工程师,后来成为了一位有影响力的风险投资家。“但一切都结束了。” 在英特尔,Andy Grove 和 Gordon Moore 得出了同样的结论。德州仪器和美国国家半导体宣布DRAM部门的亏损和裁员。《纽约时报》宣称,美国芯片行业的未来是“严峻的”。所以 Simplot 直接加入了。

America’s technology titans thought the Idaho country bumpkins didn’t have a clue. “I’d hate to say it’s over in memory chips,” said L. J. Sevin, a former Texas Instruments engineer who’d become an influential venture capitalist. “But it’s over.” At Intel, Andy Grove and Gordon Moore had reached the same conclusion. Texas Instruments and National Semiconductor announced losses and layoffs in their DRAM divisions. The future of the U.S. chip industry, the New York Times declared, was “grim.” So Simplot dove right in.

帕金森兄弟夸大了他们偏远地区的形象,讲述了漫长而曲折的故事,带有轻微的乡村拖音。事实上,他们就像作为任何硅谷初创公司的创始人,他都是老练的。两人都曾在纽约哥伦比亚大学学习,之后乔担任公司律师,而沃德则在莫斯特克设计芯片。但他们接受了他们在爱达荷州的局外人形象。他们的商业模式是席卷美国最大的芯片公司正在放弃的市场,所以他们无论如何也不会在硅谷交到很多朋友,硅谷仍在与日本的 DRAM 之战中舔舐伤口。

The Parkinson brothers played up their backcountry image, telling long, winding stories with a slight country drawl. In fact, they were as sophisticated as the founder of any Silicon Valley startup. Both had studied at Columbia University in New York, after which Joe worked as a corporate lawyer, while Ward designed chips at Mostek. But they embraced their Idaho outsider image. Their business model was to sweep into a market that America’s biggest chip firms were abandoning, so they weren’t going to make many friends anyway in Silicon Valley, which was still licking its wounds from the DRAM battles with Japan.

起初,美光嘲笑硅谷为确保政府帮助对抗日本人所做的努力。该公司假装拒绝加入由 Bob Noyce、Jerry Sanders 和 Charlie Sporck 发起的游说团体半导体行业协会。“我很清楚他们有不同的议程,”乔帕金森宣称。“他们的策略是,无论日本人进入,让我们出去。在 SIA 中占主导地位的人并没有与日本人较量。在我看来,这是弄巧成拙的策略。”

At first, Micron mocked Silicon Valley’s efforts to secure government help against the Japanese. The company sanctimoniously declined to join the Semiconductor Industry Association, the lobby group started by Bob Noyce, Jerry Sanders, and Charlie Sporck. “It was very clear to me that they had a different agenda,” Joe Parkinson declared. “Their strategy was, whatever the Japanese get into, let’s get out. The people who are dominant in the SIA are not taking the Japanese on. In my opinion, it’s a self-defeating strategy.”

美光决定在自己的游戏中挑战日本 DRAM 制造商,但要通过大幅削减成本来做到这一点。很快,该公司意识到关税可能会有所帮助,并扭转了局面,率先对进口日本 DRAM 芯片征收关税。他们指责日本生产商以低于成本的价格在美国“倾销”芯片,损害了美国生产商的利益。辛普洛对日本的贸易政策损害了他的马铃薯销售和他的记忆芯片感到愤怒。“他们对土豆征收高额关税,”他抱怨道。“我们正在为土豆付出代价。我们可以超越他们的技术,也可以超越他们的生产。我们会打败他们的。但他们正在放弃这些筹码。” 这就是他要求政府征收关税的原因。“你问我们为什么要去政府?因为法律规定他们不能这样做。”

Micron decided to challenge the Japanese DRAM makers at their own game, but to do so by aggressively cutting costs. Soon the company realized that tariffs might help, and reversed course, leading the charge for tariffs on imported Japanese DRAM chips. They accused Japanese producers of “dumping” chips in the U.S. below cost, harming American producers. Simplot was furious about Japan’s trade policies hurting his potato sales and his memory chips. “They’ve got a big tariff on potatoes,” he grumbled. “We’re paying through the nose on potatoes. We can out-tech ’em and we can out produce ’em. We’ll beat the hell out of ’em. But they’re giving those chips away.” That’s why he was demanding the government impose tariffs. “You ask why we go to the government? Cuz the law says they can’t do that.”

来自 Simplot 的关于日本公司降价过多的指控有点丰富。无论是土豆还是半导体,他总是说商业成功需要成为“最高质量产品的最低成本生产商”。无论如何,美光在削减成本方面有其硅谷或日本竞争对手无法比拟的诀窍。沃德·帕金森——“背后的工程大脑组织,”一位早期员工回忆道——拥有尽可能高效地设计 DRAM 芯片的天赋。虽然他的大多数竞争对手都专注于缩小每个芯片上晶体管和电容器的尺寸,但 Ward 意识到,如果他缩小芯片本身的尺寸,美光可以在其处理的每个圆形硅晶片上放置更多芯片。这使得制造效率大大提高。“这是迄今为止市场上最糟糕的产品,”沃德开玩笑说,“但是迄今为止生产成本最低。”

The allegation that Japanese firms were cutting prices by too much was a bit rich coming from Simplot. Whether spud or semiconductor, he’d always said business success required being “the lowest-cost producer of the highest-quality product.” Anyway, Micron had a knack for cost cuts that none of its Silicon Valley or Japanese competitors could match. Ward Parkinson—“the engineering brains behind the organization,” one early employee remembered—had a talent for designing DRAM chips as efficiently as possible. While most of his competitors were fixated on shrinking the size of transistors and capacitors on each chip, Ward realized that if he shrunk the size of the chip itself, Micron could put more chips on each of the circular silicon wafers that it processed. This made manufacturing far more efficient. “It was by far the worst product on the market,” Ward joked, “but by far the least expensive to produce.”

接下来,帕金森和他的副手简化了制造流程。制造步骤越多,每个芯片的制造时间就越多,出错的空间就越大。到 1980 年代中期,美光使用的生产步骤远远少于其竞争对手,让公司使用的设备更少,进一步降低了成本。他们调整了从 Perkin Elmer 和 ASML 购买的光刻机,使其比制造商自己认为的更精确。对炉子进行了修改,每次装载 250 个硅片,而不是行业标准的 150 个硅片。可以处理更多晶圆或减少生产时间的制造过程的每一步都意味着更低的价格。一位早期员工解释说:“我们是在动态解决这个问题的,因此与其他芯片制造商不同,我们准备做一些事情以前没有写在论文上。” 与任何日本或美国竞争对手相比,美光员工的工程专业知识更倾向于削减成本。

Next, Parkinson and his lieutenants simplified the manufacturing processes. The more steps in manufacturing, the more time each chip took to make and the more room for errors. By the mid-1980s, Micron used far fewer production steps than its competitors, letting the company use less equipment, cutting costs further. They tweaked the lithography machines they bought from Perkin Elmer and ASML to make them more accurate than the manufacturers themselves thought possible. Furnaces were modified to bake 250 silicon wafers per load rather than the 150 wafers that was industry standard. Every step of the fabrication process that could handle more wafers or reduce production times meant lower prices. “We were figuring it out on the fly,” one early employee explained, so unlike other chipmakers, “we were prepared to do things that hadn’t been written in a paper before.” More than any of its Japanese or American competitors, the engineering expertise of Micron’s employees was directed toward cost cuts.

美光无情地专注于成本,因为它别无选择。爱达荷州的一家小型初创公司根本没有其他方法可以赢得客户。这有助于博伊西的土地和电力比加利福尼亚或日本便宜,部分归功于低成本的水力发电。生存仍然是一场斗争。有一次,在 1981 年,公司的现金余额降到了如此之低,以至于只能支付两周的工资。美光勉强度过了那场危机,但几年后又一次低迷,它不得不解雇一半的员工,剩下的人减薪。自从公司成立之初,乔·帕金森就确保员工意识到他们的生存取决于效率,甚至在晚上 DRAM 价格下跌时调暗走廊灯以节省电费。员工们认为他“疯狂地”关注成本——事实证明了这一点。

Micron focused ruthlessly on costs because it had no choice. There was simply no other way for a small Idaho startup to win customers. It helped that land and electricity were cheaper in Boise than in California or in Japan, thanks in part to low-cost hydroelectric power. Survival was still a struggle. At one point, in 1981, the company’s cash balances fell so low it could cover only two weeks of payroll. Micron scraped through that crisis, but amid another downturn a few years later it had to lay off half of its employees and cut salaries for the remainder. Since the earliest days of the business, Joe Parkinson had made sure employees realized that their survival depended on efficiency, going so far as to dim hallway lights at night to save on power bills when DRAM prices fell. Employees thought he was “maniacally” focused on costs—and it showed.

美光的员工别无选择,只能维持公司的活力。在硅谷,如果你的雇主破产了,你可以沿着 101 号公路前往下一家芯片公司或计算机制造商。相比之下,美光在博伊西。“我们没有别的事可做,”一名员工解释道。“我们要么制造 DRAM,要么游戏结束。” 这是一种“勤奋的蓝领职业道德”,另一个人记得,这是一种“血汗工厂的心态”。“内存芯片是一项残酷、残酷的业务,”一位在一系列痛苦的 DRAM 市场低迷中幸存下来的早期员工回忆道。

Micron’s employees had no choice but to keep the company alive. In Silicon Valley, if your employer went bust you could drive down Route 101 to the next chip firm or computer maker. Micron, by contrast, was in Boise. “We didn’t have something else to do,” one employee explained. “We either made DRAMs, or game over.” It was a “hardworking, blue collar work ethic,” another remembered, a “sweatshop mentality.” “Memory chips is a brutal, brutal business,” recalled an early employee who survived a series of painful DRAM market downturns.

杰克辛普洛特从未失去信心。他在他曾经拥有的每一项业务中都度过了低迷时期。他不会因为短期价格波动而放弃美光。尽管在日本竞争达到顶峰时进入 DRAM 市场,但美光幸存下来并最终蓬勃发展。大多数其他美国 DRAM 生产商在 1980 年代后期被迫退出市场。TI 继续生产 DRAM 芯片,但难以赚钱,最终将其业务出售给了美光。Simplot 的第一笔 100 万美元投资最终激增为 10 亿美元的股份。

Jack Simplot never lost faith. He’d survived downswings in every business he’d ever owned. He wasn’t going to abandon Micron because of short-term price swings. Despite entering the DRAM market just as Japanese competition was peaking, Micron survived and eventually thrived. Most other American DRAM producers were forced out of the market in the late 1980s. TI kept manufacturing DRAM chips but struggled to make any money, and eventually sold its operations to Micron. Simplot’s first $1 million investment eventually ballooned into a billion-dollar stake.

美光学会了在每一代 DRAM 芯片的存储容量方面与东芝和富士通等日本竞争对手竞争,并在成本上超越它们。与 DRAM 行业的其他公司一样,美光的工程师们打破了物理定律,制造了更密集的 DRAM 芯片,提供了个人电脑所需的内存芯片。但仅靠先进技术并不足以拯救美国的 DRAM 产业。英特尔和 TI 拥有大量技术,但无法让业务运转起来。美光爱达荷州好斗的工程师凭借他们的创造力和削减成本的技巧在太平洋两岸战胜了竞争对手。经过十年的痛苦,美国芯片行业终于取得了胜利——这要归功于美国最伟大的马铃薯种植者的市场智慧。

Micron learned to compete with Japanese rivals like Toshiba and Fujitsu when it came to the storage capacity of each generation of DRAM chip and to outcompete them on cost. Like the rest of the DRAM industry, Micron’s engineers bent the laws of physics as they made ever denser DRAM chips, providing the memory chips needed in personal computers. But advanced technology on its own wasn’t enough to save America’s DRAM industry. Intel and TI had plenty of technology but couldn’t make the business work. Micron’s scrappy Idaho engineers outmaneuvered rivals on both sides of the Pacific with their creativity and cost-cutting skill. After a decade of pain, the U.S. chip industry finally scored a win—and it was only possible thanks to the market wisdom of America’s greatest potato farmer.

第 22 章 颠覆英特尔

CHAPTER 22 Disrupting Intel

看,克莱顿我是一个忙碌的人,我没有时间阅读学者们的胡言乱语,”安迪·格罗夫告诉哈佛商学院最著名的教授克莱顿·克里斯滕森。几年后,当他们两人登上《福布斯》的封面时,克里斯滕森——六英尺八英寸高——耸立在格罗夫的上方,而格罗夫的秃头几乎没有到达克里斯滕森的肩膀。但是格鲁夫的强度超过了他周围的每个人。他是一个“踢匈牙利人,”他的长期副手解释说,“咀嚼人们的脚踝,对他们大喊大叫,挑战他们,并尽可能用力推动。” 最重要的是,格鲁夫的坚韧使英特尔免于破产,并使其成为世界上最赚钱和最强大的公司之一。

“Look, Clayton, I’m a busy man and I don’t have time to read drivel from academics,” Andy Grove told Harvard Business School’s most famous professor, Clayton Christensen. When the two of them made the cover of Forbes several years later, Christensen—six feet, eight inches tall—towered over Grove, whose balding head barely reached Christensen’s shoulder. But Grove’s intensity outshone that of everyone around him. He was a “butt-kicking Hungarian,” his longtime deputy explained, “chewing on people’s ankles, and yelling at them, and challenging them, and pushing as hard as he could.” More than anything else did, Grove’s tenacity saved Intel from bankruptcy and made it one of the world’s most profitable and powerful companies.

克里斯蒂安森教授以其“颠覆性创新”理论而闻名,其中一项新技术取代了现有公司。随着 DRAM 业务的下滑,格鲁夫意识到英特尔——曾经是创新的代名词——现在正在被颠覆。到 1980 年代初,格鲁夫已成为英特尔的总裁,负责日常运营,尽管摩尔仍然发挥着重要作用。格鲁夫在他的畅销书《只有偏执狂才能生存》中描述了他的管理理念:“对竞争的恐惧、对破产的恐惧、对错误的恐惧和对失败的恐惧都可以成为强大的动力。” 经过一整天的工作,恐惧让格鲁夫不停地翻阅他的信件或与下属的电话,担心他会错过产品延迟或客户不满意的消息。从表面上看,安迪·格罗夫实现了美国梦:一个曾经一贫如洗的难民变成了科技巨头。在这个硅谷的成功故事中,有一个匈牙利流亡者,童年时期曾躲避在布达佩斯街头行进的苏联和纳粹军队,这让他留下了伤疤。

Professor Christiansen was famous for his theory of “disruptive innovation,” in which a new technology displaces incumbent firms. As the DRAM business slumped, Grove realized that Intel—once synonymous with innovation—was now being disrupted. By the early 1980s, Grove was Intel’s president, in charge of day-to-day operations, though Moore still played a major role. Grove described his management philosophy in his bestselling book Only the Paranoid Survive: “Fear of competition, fear of bankruptcy, fear of being wrong and fear of losing can all be powerful motivators.” After a long day of work, it was fear that kept Grove flipping through his correspondence or on the phone with subordinates, worried he’d missed news of product delays or unhappy customers. On the outside, Andy Grove was living the American dream: a once-destitute refugee transformed into a tech titan. Inside this Silicon Valley success story was a Hungarian exile scarred by a childhood spent hiding from the Soviet and Nazi armies marching down Budapest streets.

Grove 意识到英特尔销售 DRAM 芯片的商业模式已经结束。DRAM 价格可能会从价格暴跌中恢复,但英特尔永远不会赢回市场份额。它已被日本生产商“破坏”。现在它要么破坏自己,要么失败。退出 DRAM 市场似乎是不可能的。英特尔开创了内存芯片,承认失败将是一种耻辱。一位员工说,这就像福特决定下车一样。“我们怎么能放弃我们的身份?” 格鲁夫想知道。1985 年的大部分时间里,他都坐在位于英特尔圣克拉拉总部的戈登摩尔办公室里,他们两人凝视着窗外远处 Great America 游乐园的摩天轮,希望就像摩天轮上的一个小木屋一样,内存市场最终会触底并重新开始盘旋。

Grove realized Intel’s business model of selling DRAM chips was finished. DRAM prices might recover from the price slump, but Intel would never win back market share. It had been “disrupted” by Japanese producers. Now it would either disrupt itself or fail. Exiting the DRAM market felt impossible. Intel had pioneered memory chips, and admitting defeat would be humiliating. It was like Ford deciding to get out of cars, one employee said. “How could we give up our identity?” Grove wondered. He spent much of 1985 sitting in Gordon Moore’s office at Intel’s Santa Clara headquarters, the two of them staring out the window at the Ferris wheel in the Great America amusement park in the distance, hoping that like one of the cabins on the Ferris wheel, the memory market would eventually hit bottom and begin circling up again.

然而,灾难性的 DRAM 数字是不可否认的。英特尔永远不会在内存方面赚到足够的钱来证明新投资的合理性。不过,它在日本公司仍然落后的小型微处理器市场上处于领先地位。该领域的一项发展提供了一线希望。1980 年,英特尔赢得了与美国计算机巨头 IBM 的一份小合同,为一种称为个人计算机的新产品制造芯片。IBM 与一位名叫比尔盖茨的年轻程序员签订合同,为计算机操作系统编写软件。1981 年 8 月 12 日,背景是华尔道夫酒店大宴会厅的华丽墙纸和厚窗帘,IBM 宣布推出其个人电脑,售价 1,565 美元,包括一台笨重的电脑、一个大屏幕显示器、一个键盘、一台打印机和两个软盘驱动器。它里面有一个小的英特尔芯片。

However, the disastrous DRAM numbers were impossible to deny. Intel would never make enough money in memory to justify new investments. It was a leader, though, in the small microprocessor market, where Japanese firms still lagged. And one development in that arena provided a glimmer of hope. In 1980, Intel had won a small contract with IBM, America’s computer giant, to build chips for a new product called a personal computer. IBM contracted with a young programmer named Bill Gates to write software for the computer’s operating system. On August 12, 1981, with the ornate wallpaper and thick drapes of the Waldorf Astoria’s grand ballroom in the background, IBM announced the launch of its personal computer, priced at $1,565 for a bulky computer, a big-box monitor, a keyboard, a printer, and two diskette drives. It had a small Intel chip inside.

微处理器市场似乎几乎肯定会增长。但微处理器销售量可能超过构成芯片销售量大头的 DRAM 的前景似乎令人难以置信,格罗夫的作品之一代表们召回。格罗夫别无选择。“如果我们被踢出局,董事会任命了一位新 CEO,你认为他会怎么做?” 格鲁夫问摩尔,他想继续生产 DRAM 芯片。“他会让我们摆脱记忆,”摩尔羞怯地承认。最后,英特尔决定离开存储器,将 DRAM 市场拱手让给日本人,专注于 PC 微处理器。对于一家建立在 DRAM 上的公司来说,这是一场大胆的赌博。“颠覆性创新”在 Clayton Christensen 的理论中听起来很有吸引力,但在实践中却令人痛心,这是一个“咬牙切齿”的时代,格罗夫回忆道,“争吵和争论。” 破坏是显而易见的。如果有的话,这项创新将需要数年时间才能得到回报。

The microprocessor market seemed almost certain to grow. But the prospect that microprocessor sales could overtake DRAMs, which constituted the bulk of chip sales, seemed mind-boggling, one of Grove’s deputies recalled. Grove saw no other choice. “If we got kicked out and the board brought in a new CEO, what do you think he would do?” Grove asked Moore, who wanted to keep producing DRAM chips. “He would get us out of memories,” Moore admitted sheepishly. Finally, Intel decided to leave memories, surrendering the DRAM market to the Japanese and focusing on microprocessors for PCs. It was a gutsy gamble for a company that had been built on DRAMs. “Disruptive innovation” sounded attractive in Clayton Christensen’s theory, but it was gut-wrenching in practice, a time of “gnashing of teeth,” Grove remembered, and “bickering and arguments.” The disruption was obvious. The innovation would take years to pay off, if it ever did.

在等着看他在个人电脑上的赌注是否会奏效时,格鲁夫以硅谷罕见的冷酷无情运用了他的偏执狂。工作日从早上 8 点开始,任何迟到的人都会受到公开批评。员工之间的分歧通过 Grove 称为“建设性对抗。” 他的副手克雷格·巴雷特(Craig Barrett)打趣说,他的管理技巧是“抓住一个人,用大锤猛击他们的头。”

While waiting to see if his bet on PCs would work, Grove applied his paranoia with a ruthlessness Silicon Valley had rarely seen. Workdays started at 8 a.m. sharp and anyone who signed in late was criticized publicly. Disagreements between employees were resolved via a tactic Grove called “constructive confrontation.” His go-to management technique, quipped his deputy Craig Barrett, was “grabbing someone and slamming them over the head with a sledgehammer.”

这不是硅谷闻名的随心所欲的文化,但英特尔需要一名训练中士。其 DRAM 芯片面临与其他美国芯片制造商相同的质量问题。当它在 DRAM 上赚钱时,它是通过率先推出新设计的市场来实现的,而不是通过成为大规模生产的领导者。Bob Noyce 和 Gordon Moore 一直专注于保持尖端技术。但诺伊斯承认,他总是觉得“冒险部分”比“冒险部分”更有趣。“控制部分。” Grove 非常喜欢控制,这就是 Gordon Moore 在 1963 年首次将他带到 Fairchild 的原因:解决公司的生产问题。当他跟随诺伊斯和摩尔来到英特尔时,他被赋予了同样的角色。格罗夫的余生都沉浸在公司制造流程和业务的每一个细节中,在一种挥之不去的恐惧感的驱使下。

This wasn’t the freewheeling culture Silicon Valley was known for, but Intel needed a drill sergeant. Its DRAM chips faced the same quality problems as those of other American chipmakers. When it had made money in DRAMs, it did so by being first to the market with a new design, not by being the leader in mass production. Bob Noyce and Gordon Moore had always fixated on maintaining cutting-edge tech. But Noyce admitted that he always found “the venture part” more fun than “the control part.” Grove loved control as much as anything, which is why Gordon Moore had first brought him to Fairchild in 1963: to solve the company’s production problems. When he followed Noyce and Moore to Intel, he was given the same role. Grove spent the rest of his life immersed in every detail of the company’s manufacturing processes and its business, driven by a nagging sense of fear.

在格鲁夫的重组计划中,第一步是裁员超过 25% 的英特尔员工,关闭位于硅谷、俄勒冈州和波多黎各的工厂里科和巴巴多斯。格罗夫的副手将他老板的做法描述为:“天啊。解雇这两个人,烧毁船只,扼杀生意。” 他以诺伊斯和摩尔永远不可能做到的方式无情和果断。第二步是制造工作。他和巴雷特无情地复制了日本的制造方法。“巴雷特基本上带着棒球 [球棒] 去制造并说:‘该死的!我们不会被日本人打败,”一名下属回忆道。他强迫工厂经理访问日本并告诉他们:“这就是你应该做的。”

In Grove’s restructuring plan, step one was to lay off over 25 percent of Intel’s workforce, shutting facilities in Silicon Valley, Oregon, Puerto Rico, and Barbados. Grove’s deputy described his boss’s approach as: “Oh my god. Fire these two people, burn the ships, kill the business.” He was ruthless and decisive in a way that Noyce and Moore never could have been. Step two was to make manufacturing work. He and Barrett relentlessly copied Japanese manufacturing methods. “Barrett basically took a baseball [bat] to manufacturing and said: ‘Damn it! We are not going to get beaten by the Japanese,’ ” one subordinate recalled. He forced factory managers to visit Japan and told them: “This is how you are supposed to do it.”

英特尔的新制造方法被称为“完全复制”。一旦英特尔确定一组特定的生产流程效果最佳,它们就会被复制到所有其他英特尔设施中。在此之前,工程师们以微调英特尔的流程而自豪。现在他们被要求不要思考,而是要复制。“这是一个巨大的文化问题,”一个人回忆道,随心所欲的硅谷风格被流水线的严谨所取代。“我被视为独裁者,”巴雷特承认。但“完全复制”奏效了:英特尔的产量大幅上升,而其制造设备的使用效率更高,从而降低了成本。公司的每个工厂开始不再像研究实验室那样运作,更像是一台经过微调的机器。

Intel’s new manufacturing method was called “copy exactly.” Once Intel determined that a specific set of production processes worked best, they were replicated in all other Intel facilities. Before then, engineers had prided themselves on fine-tuning Intel’s processes. Now they were asked not to think, but to replicate. “It was a huge cultural issue,” one remembered, as a freewheeling Silicon Valley style was replaced with assembly line rigor. “I was perceived as a dictator,” Barrett admitted. But “copy exactly” worked: Intel’s yields rose substantially, while its manufacturing equipment was used more efficiently, driving down costs. Each of the company’s plants began to function less like a research lab and more like a finely tuned machine.

格鲁夫和英特尔也很幸运。一些在 1980 年代初有利于日本生产商的结构性因素开始发生变化。1985 年至 1988 年间,日元对美元的价值翻了一番,使美国的出口商品更便宜。美国的利率在 1980 年代急剧下降,降低了英特尔的资本成本。与此同时,总部位于得克萨斯州的康柏电脑在 IBM 的 PC 市场上大放异彩,其驱动力是意识到虽然编写操作系统或制造微处理器很困难,但将 PC 组件组装到塑料盒中相对简单。康柏推出了使用英特尔芯片和微软软件的自己的个人电脑,价格远低于 IBM 的个人电脑。到 1980 年代中期,康柏和其他公司正在构建 IBM 个人电脑的“克隆”销量超过 IBM 本身。随着每个办公室和许多家庭都安装了计算机,价格急剧下降。除了苹果电脑,几乎每台 PC 都使用英特尔的芯片和 Windows 软件,这两种软件的设计目的都是为了顺畅地协同工作。英特尔进入了个人电脑时代,几乎垄断了个人电脑的芯片销售。

Grove and Intel got lucky, too. Some of the structural factors that had favored Japanese producers in the early 1980s began to shift. Between 1985 and 1988, the value of the Japanese yen doubled against the dollar, making American exports cheaper. Interest rates in the U.S. fell sharply over the 1980s, reducing Intel’s capital costs. Meanwhile, Texas-based Compaq Computer muscled in on IBM’s PC market, driven by the realization that though it was hard to write operating systems or build microprocessors, assembling PC components into a plastic box was relatively straightforward. Compaq launched its own PCs using Intel chips and Microsoft software, priced far below IBM’s PCs. By the mid-1980s, Compaq and other firms building “clones” of IBM’s PC sold more units than IBM itself. Prices fell precipitously as computers were installed in every office and many homes. Except for Apple’s computers, almost every PC used Intel’s chips and Windows software, both of which had been designed to work smoothly together. Intel entered the personal computer era with a virtual monopoly on chip sales for PCs.

格鲁夫对英特尔的重组是硅谷资本主义的教科书案例。他认识到该公司的商业模式已被打破,并决定放弃英特尔成立时要制造的 DRAM 芯片,从而“颠覆”英特尔本人。该公司在 PC 芯片市场上建立了垄断地位,每两年发布新一代芯片,提供更小的晶体管和更多的处理能力。只有偏执狂才能生存,安迪·格罗夫相信。拯救英特尔的不仅仅是创新或专业知识,而是他的偏执狂。

Grove’s restructuring of Intel was a textbook case of Silicon Valley capitalism. He recognized that the company’s business model was broken and decided to “disrupt” Intel himself by abandoning the DRAM chips it had been founded to build. The firm established a stranglehold on the market for PC chips, issuing a new generation of chip every year or two, offering smaller transistors and more processing power. Only the paranoid survive, Andy Grove believed. More than innovation or expertise, it was his paranoia that saved Intel.

第 23 章 “我的敌人的敌人”:韩国的崛起

CHAPTER 23 “My Enemy’s Enemy”: The Rise of Korea

李秉哲几乎可以卖任何东西来获利。李出生于 1910 年,就在杰克辛普洛之后一年,他于 1938 年 3 月开始了他的商业生涯,当时他的祖国韩国是日本帝国的一部分,与中国交战,很快又与美国交战。李的第一批产品是干鱼和蔬菜,他从韩国收集并运往中国北方,为日本的战争机器提供食物。韩国是一个贫穷的死水,没有工业和技术,但李已经梦想着建立一个能够成为“大、强、永恒,”他宣称。由于两个有影响力的盟友:美国的芯片行业和韩国政府,他将三星变成了一个半导体超级大国。硅谷战胜日本人的战略的一个关键部分是在亚洲寻找更便宜的供应来源。Lee 认为这是三星可以轻松扮演的角色。

Lee Byung-Chul could make a profit selling almost anything. Born in 1910, just a year after Jack Simplot, Lee launched his business career in March 1938, a time when his native Korea was part of Japan’s empire, at war with China and soon with the United States. Lee’s first products were dried fish and vegetables, which he gathered from Korea and shipped to northern China to feed Japan’s war machine. Korea was an impoverished backwater, with no industry or technology, but Lee was already dreaming of building a business that would be “big, strong, and eternal,” he declared. He would turn Samsung into a semiconductor superpower thanks to two influential allies: America’s chip industry and the South Korean state. A key part of Silicon Valley’s strategy to outmaneuver the Japanese was to find cheaper sources of supply in Asia. Lee decided this was a role Samsung could easily play.

韩国习惯于在更大的竞争对手之间穿梭。在李创立三星七年后,它本可以在 1945 年日本被美国击败后被粉碎。然而,李灵巧地转过身来,像兜售干鱼一样顺利地交易政治赞助人。他与战后占领韩国南半部的美国人建立了联系,并抵御了想要破坏的韩国政客像他这样的大企业集团。当朝鲜共产党政府入侵南方时,他甚至保留了自己的资产——然而,当敌人短暂占领首尔时,一名共产党总书记夺取了李的雪佛兰和把它开到被占领的首都周围。

South Korea was used to navigating between bigger rivals. Seven years after Lee founded Samsung, it could have been crushed in 1945, following Japan’s defeat by the United States. Yet Lee deftly pivoted, trading political patrons as smoothly as he hawked dried fish. He forged ties with the Americans who occupied the southern half of Korea after the war and fended off South Korean politicians who wanted to break up big business groups like his. He even kept hold of his assets when the Communist government in North Korea invaded the South—though, when the enemy briefly captured Seoul, a Communist Party chief seized Lee’s Chevrolet and drove it around the occupied capital.

尽管发生了战争,李还是扩大了他的商业帝国,巧妙地驾驭了韩国复杂的政治。1961 年军政府掌权后,将军们剥夺了李的银行,但他的其他公司完好无损地幸存了下来。他坚称三星是在为国家的利益而努力——国家的利益取决于三星成为一家世界级的公司。“通过商业服务国家,”李氏家族格言的第一部分是这样写的。从鱼和蔬菜开始,他涉足制糖、纺织、化肥、建筑、银行和保险。他将 1960 年代和 1970 年代韩国的经济繁荣视为他为国家服务的证明。批评者指出,到 1960 年,他已成为韩国首富,他们认为他的财富证明了这个国家——以及它的腐败政客——正在为他服务。

Lee expanded his business empire despite the war, navigating South Korea’s complicated politics with finesse. When a military regime took power in 1961, the generals stripped Lee of his banks, but he survived with his other companies intact. He insisted Samsung was working for the good of the nation—and that the good of the nation depended on Samsung becoming a world-class company. “Serving the nation through business,” the first part of the Lee family motto read. From fish and vegetables, he diversified into sugar, textiles, fertilizer, construction, banking, and insurance. He saw Korea’s economic boom during the 1960s and 1970s as proof he was serving the nation. Critics, who noted that by 1960 he had become the richest person in South Korea, thought his wealth was evidence the nation—and its venal politicians—were serving him.

Lee 长期以来一直想打入半导体行业,看着东芝和富士通等公司在 1970 年代末和 1980 年代初占据 DRAM 市场份额。韩国已经是美国或日本制造的芯片外包组装和包装的重要地点。此外,美国政府在 1966 年帮助创建了韩国科学技术学院,越来越多的韩国人从美国顶尖大学毕业或在韩国接受美国教育的教授的培训。然而,即使拥有熟练的劳动力,公司也不容易从基本组装跳到尖端芯片制造。三星此前曾涉足简单的半导体工作,但努力赚钱或生产先进技术。

Lee had long wanted to break into the semiconductor industry, watching companies like Toshiba and Fujitsu take DRAM market share in the late 1970s and early 1980s. South Korea was already an important location for outsourced assembly and packaging of chips made in the U.S. or Japan. Moreover, the U.S. government had helped fund the creation in 1966 of the Korea Institute of Science and Technology, and a growing number of Koreans were graduating from top U.S. universities or being trained in Korea by U.S.-educated professors. Even with a skilled workforce, though, it wasn’t easy for firms to jump from basic assembly to cutting-edge chipmaking. Samsung had previously dabbled in simple semiconductor work but struggled to make money or produce advanced technology.

然而,在 1980 年代初期,李感觉到环境的变化。1980 年代硅谷和日本之间残酷的 DRAM 竞争提供了一个机会。与此同时,韩国政府已将半导体确定为优先事项。作为李考虑到三星的未来,他于 1982 年春天前往加利福尼亚,参观了惠普的设施,并惊叹于该公司的技术。如果惠普可以从帕洛阿尔托的车库发展成为科技巨头,那么像三星这样的鱼和蔬菜店肯定也可以。“这一切都归功于半导体,”一位惠普员工告诉他。他还参观了一家 IBM 电脑工厂,并为自己被允许拍照感到震惊。“你的工厂一定有很多秘密,”他告诉带他参观的 IBM 员工。“仅凭观察是无法复制的,”该员工自信地回答。然而,复制硅谷的成功正是李计划要做的。

In the early 1980s, however, Lee sensed the environment changing. The brutal DRAM competition between Silicon Valley and Japan during the 1980s provided an opening. The South Korean government, meanwhile, had identified semiconductors as a priority. As Lee pondered Samsung’s future, he traveled to California in spring 1982, visiting Hewlett-Packard’s facilities and marveling at the company’s technology. If HP could grow from a Palo Alto garage to a tech behemoth, surely a fish-and-vegetables shop like Samsung could, too. “It’s all thanks to semiconductors,” one HP employee told him. He also toured an IBM computer factory and was shocked he was allowed to take photographs. “There must be many secrets in your factory,” he told the IBM employee giving him the tour. “They can’t be replicated by mere observation,” the employee confidently responded. Replicating Silicon Valley’s success, though, was exactly what Lee planned to do.

这样做需要数百万美元的资本支出,但不能保证它会奏效。即使对李来说,这也是一个很大的赌注。他犹豫了几个月。失败可能会摧毁他的整个商业帝国。然而,韩国政府表示愿意提供财政支持。它曾承诺投资 4 亿美元发展其半导体产业。韩国的银行将听从政府的指示,再贷出数百万美元。因此,与日本一样,韩国的科技公司并非来自车库,而是来自能够获得廉价银行贷款和政府支持的大型企业集团。1983 年 2 月,经过一个紧张不安的不眠之夜,李拿起电话,给三星电子部门负责人打了电话,宣称:“三星将制造半导体。” 他把公司的未来押在半导体上,他宣称,至少 1 亿美元。

Doing so would require many millions of dollars in capital expenditure, yet there was no guarantee it would work. Even for Lee, this was a big bet. He hesitated for months. Failure could bring down his entire business empire. South Korea’s government, however, signaled it was willing to provide financial support. It had promised to invest $400 million to develop its semiconductor industry. Korea’s banks would follow the government’s direction and lend millions more. As in Japan, therefore, Korea’s tech companies emerged not from garages, but from massive conglomerates with access to cheap bank loans and government support. In February 1983, after a nervous, sleepless night, Lee picked up the phone, called the head of Samsung’s electronics division, and proclaimed: “Samsung will make semiconductors.” He bet the company’s future on semiconductors, and was ready to spend at least $100 million, he declared.

李是一位精明的企业家,韩国政府坚定地支持他。然而,如果没有硅谷的支持,三星在芯片上的全部赌注是不会奏效的。硅谷认为,应对来自日本的内存芯片国际竞争的最佳方法是在韩国寻找更便宜的来源,同时将美国的研发工作重点放在价值更高的产品上,而不是商品化的 DRAM。因此,美国芯片制造商将韩国新贵视为潜在合作伙伴。“有韩国人在身边,”鲍勃·诺伊斯告诉安迪·格罗夫,日本的“不惜一切代价倾销”战略不会成功垄断世界DRAM生产,因为韩国人会削弱日本生产商。诺伊斯预测,结果对日本芯片制造商来说将是“致命的”。

Lee was a canny entrepreneur, and South Korea’s government stood firmly behind him. Yet Samsung’s all-in bet on chips wouldn’t have worked without support from Silicon Valley. The best way to deal with international competition in memory chips from Japan, Silicon Valley wagered, was to find an even cheaper source in Korea, while focusing America’s R&D efforts on higher-value products rather than commoditized DRAMs. U.S. chipmakers therefore saw Korean upstarts as potential partners. “With the Koreans around,” Bob Noyce told Andy Grove, Japan’s strategy of “dump no matter what the costs” wouldn’t succeed in monopolizing the world’s DRAM production, because the Koreans would undercut Japanese producers. The result would be “deadly” to Japanese chipmakers, Noyce predicted.

英特尔因此为韩国 DRAM 生产商的崛起而欢呼。它是 1980 年代与三星签署合资企业的几家硅谷公司之一,销售三星以英特尔自有品牌制造的芯片,并押注帮助韩国芯片产业将减少日本对硅谷的威胁。此外,韩国的成本和工资远低于日本,因此三星等韩国公司即使制造工艺不如高效的日本公司那样完美调整,也有机会赢得市场份额。

Intel therefore cheered the rise of Korean DRAM producers. It was one of several Silicon Valley firms to sign a joint venture with Samsung in the 1980s, selling chips Samsung manufactured under Intel’s own brand and wagering that helping Korea’s chip industry would reduce Japan’s threat to Silicon Valley. Moreover, Korea’s costs and wages were substantially lower than Japan’s, so Korean firms like Samsung had a shot at winning market share even if their manufacturing processes weren’t as perfectly tuned as the ultra-efficient Japanese.

美日贸易紧张局势也帮助了韩国公司。1986 年,华盛顿威胁要征收关税,除非日本停止“倾销”——在美国市场低价出售 DRAM 芯片——东京同意限制其对美国的芯片销售,并承诺不会低价出售。这为韩国公司提供了以更高价格销售更多 DRAM 芯片的机会。美国人不打算让这笔交易使韩国公司受益,但他们很高兴看到除了日本以外的任何人都生产他们需要的芯片。

U.S.-Japan trade tension helped Korean companies, too. After Washington threatened tariffs unless Japan stopped “dumping”—selling DRAM chips cheaply on the U.S. market—in 1986, Tokyo agreed to limit its sales of chips to the U.S. and promised not to sell at low prices. This provided an opening for Korean companies to sell more DRAM chips at higher prices. The Americans didn’t intend for the deal to benefit Korean firms, but they were happy to see anyone but Japan producing the chips they needed.

美国不仅为韩国 DRAM 芯片提供了市场;它也提供了技术。由于硅谷的 DRAM 生产商大多濒临倒闭,因此毫不犹豫地将一流技术转移到韩国。Lee 提议从资金紧张的存储芯片初创公司美光公司获得 64K DRAM 的设计许可,并在此过程中与其创始人 Ward Parkinson 成为朋友。爱达荷人正在寻找他们能得到的任何钱,他们热切地同意了,即使这意味着三星会学习他们的许多流程。“无论我们做什么,三星都做到了,”帕金森回忆道,他认为三星提供的现金注入对于帮助美光生存来说“并不重要,但很接近”。一些行业领导者,比如戈登·摩尔,担心一些芯片公司如此绝望,以至于他们会“放弃越来越有价值的技术”。然而,内存芯片几乎破产。硅谷的大部分人都乐于与韩国公司合作,削弱日本竞争对手,并帮助韩国成为世界领先的存储芯片制造中心之一。正如杰里桑德斯解释的那样,逻辑很简单:“我敌人的敌人就是我的朋友。”

The U.S. didn’t simply provide a market for South Korean DRAM chips; it provided technology, too. With Silicon Valley’s DRAM producers mostly near collapse, there was little hesitation about transferring top-notch technology to Korea. Lee proposed to license a design for a 64K DRAM from Micron, the cash-strapped memory chip startup, befriending its founder Ward Parkinson in the process. The Idahoans, looking for any money they could get, eagerly agreed even if it meant Samsung would learn many of their processes. “Whatever we did, Samsung did,” Parkinson remembered, seeing the cash infusion that Samsung provided as “not crucial, but close” in helping Micron survive. Some industry leaders, like Gordon Moore, worried that some chip firms were so desperate they’d “part with increasingly valuable bits of technology.” However, it was hard to make the case that DRAM technology was particularly valuable when most U.S. firms making memory chips were nearly bankrupt. Most of Silicon Valley was happy to work with Korean companies, undercutting Japanese competitors and helping make South Korea one of the world’s leading centers of memory chipmaking. The logic was simple, as Jerry Sanders explained: “my enemy’s enemy is my friend.”

第 24 章 “这就是未来”

CHAPTER 24 “This Is the Future”

在日本 DRAM 的猛攻之后,美国芯片业的重生只有安迪·格鲁夫的偏执狂、杰里·桑德斯的赤脚斗殴和杰克·辛普洛特的牛仔竞争力,才有可能重生。硅谷由睾酮和股票期权推动的竞争往往不像教科书中描述的枯燥的经济学,更像是达尔文式的适者生存的斗争。许多公司倒闭,财富流失,数以万计的员工被解雇。像英特尔和美光这样的公司幸存下来,与其说是因为他们的工程技能——尽管这些很重要——不如说是因为他们能够利用技术才能在一个竞争激烈、无情的行业中赚钱。

The rebirth of America’s chip industry after Japan’s DRAM onslaught was only possible thanks to Andy Grove’s paranoia, Jerry Sanders’s bare-knuckle brawling, and Jack Simplot’s cowboy competitiveness. Silicon Valley’s testosterone and stock option−fueled competition often felt less like the sterile economics described in textbooks and more like a Darwinian struggle for the survival of the fittest. Many firms failed, fortunes were lost, and tens of thousands of employees were laid off. The companies like Intel and Micron that survived did so less thanks to their engineering skills—though these were important—than their ability to capitalize on technical aptitude to make money in a hypercompetitive, unforgiving industry.

然而,硅谷的重生不仅仅是英雄企业家和创造性破坏的故事。随着这些新工业巨头的崛起,一批新的科学家和工程师正在准备芯片制造的飞跃,并设计出革命性的新方法来使用处理能力。许多这些发展是与政府的努力相协调的,通常不是国会或白宫的强权,而是像 DARPA 这样的小型、灵活的组织的工作,这些组织有权对未来技术进行大赌注——并建立教育和此类赌博所需的研发基础设施。

Yet Silicon Valley’s rebirth isn’t solely a story of heroic entrepreneurs and creative destruction. Alongside the rise of these new industrial titans, a new set of scientists and engineers were preparing a leap forward in chipmaking and devising revolutionary new ways to use processing power. Many of these developments occurred in coordination with government efforts, usually not the heavy hand of Congress or the White House, but the work of small, nimble organizations like DARPA that were empowered to take big bets on futuristic technologies—and to build the educational and R&D infrastructure that such gambles required.

来自日本高质量、低成本 DRAM 芯片的竞争并不是硅谷在 1980 年代面临的唯一问题。Gordon Moore 的著名定律预测每个芯片上的晶体管数量会呈指数增长,但这个梦想变得越来越难以实现。到 1970 年代后期,许多集成电路的设计过程与英特尔的 Federico Faggin 用于生产第一个微处理器的过程相同。1971 年,Faggin 花了半年时间蹲在他的绘图桌前,用英特尔最先进的工具勾勒出设计草图:直尺和彩色铅笔。然后,用小刀将这个设计切割成红色胶片 Rubylith。一台特殊的相机将 Rubylith 雕刻的图案投射到面具上,这是一块镀铬覆盖的玻璃板,完美地复制了 Rubylith 的图案。最后,光线通过掩模和一组透镜照射,将微小版本的图案投射到硅片上。经过几个月的素描和雕刻,Faggin 创造了一个芯片。

Competition from Japan’s high-quality, low-cost DRAM chips wasn’t the only problem Silicon Valley faced in the 1980s. Gordon Moore’s famous law predicted exponential growth in the number of transistors on each chip, but this dream was getting ever more difficult to realize. Through the late 1970s, many integrated circuits had been designed by the same process Intel’s Federico Faggin used to produce the first microprocessor. In 1971, Faggin had spent half a year crouched over his drafting table, sketching the design with Intel’s most advanced tools: a straightedge and color pencils. Then, this design was cut into Rubylith, a red film, using a penknife. A special camera projected the patterns carved in Rubylith onto a mask, a glass plate with a chrome covering that perfectly replicated the Rubylith’s pattern. Finally, light was shined through the mask and a set of lenses to project a tiny version of the pattern on a silicon wafer. After months of sketching and carving, Faggin had created a chip.

问题是,虽然铅笔和镊子对于具有一千个组件的集成电路来说已经足够了,但对于具有一百万个晶体管的芯片来说,则需要更复杂的工具。Carver Mead 是戈登·摩尔的朋友,留着山羊胡子的物理学家,当他被介绍给施乐帕洛阿尔托研究中心的计算机架构师 Lynn Conway 时,他正在为这个困境感到困惑,当时刚刚发明了带有鼠标和键盘的个人计算机的概念。

The problem was, while pencils and tweezers were adequate tools for an integrated circuit with a thousand components, something more sophisticated was needed for a chip with a million transistors. Carver Mead, the goateed physicist who was a friend of Gordon Moore, was puzzling over this dilemma when he was introduced to Lynn Conway, a computer architect at Xerox’s Palo Alto Research Center, where the concept of the personal computer with a mouse and a keyboard was just then being invented.

康威是一位才华横溢的计算机科学家,但任何与她交谈的人都会发现她的头脑中闪耀着来自不同领域的见解,从天文学到人类学再到历史哲学。她于 1973 年抵达施乐公司“隐身模式,”她解释说,在经历了性别转变后,她于 1968 年被 IBM 解雇。她震惊地发现硅谷的芯片制造商更像是艺术家而不是工程师。高科技工具与简单的镊子配对。芯片制造商在每块硅片上制作出极其复杂的图案,但他们的设计方法是中世纪工匠的方法。每个公司的晶圆厂(制造厂)都有一个漫长而复杂的,如果要在该特定设施中生产芯片,必须如何设计芯片的专有指令集。康威作为计算机架构师的培训教会了她根据构建任何计算机程序的标准化指令进行思考,她发现了这种方法奇怪的落后。

Conway was a brilliant computer scientist, but anyone who spoke with her discovered a mind that glistened with insights from diverse fields, astronomy to anthropology to historical philosophy. She had arrived at Xerox in 1973 in “stealth mode,” she explained, following being fired from IBM in 1968 after undergoing a gender transition. She was shocked to find that the Valley’s chipmakers were more like artists than engineers. High-tech tools were paired with simple tweezers. Chipmakers produced marvelously complex patterns on each block of silicon, but their design methods were those of medieval artisans. Each company’s fab (fabrication plant) had a long, complicated, proprietary set of instructions for how chips must be designed if they were to be produced in that specific facility. Conway, whose training as a computer architect had taught her to think in terms of the standardized instructions on which any computer program is built, found this method bizarrely backward.

康威意识到,米德预言的数字革命需要严格的算法。在她和 Mead 被一个共同的同事介绍后,他们开始讨论如何标准化芯片设计。他们想知道,为什么你不能对机器进行编程来设计电路。“一旦你可以编写一个程序来做某事,”米德宣称,“你就不需要任何人的工具箱,你自己写。”

Conway realized that the digital revolution Mead prophesied needed algorithmic rigor. After she and Mead were introduced by a mutual colleague, they began discussing how to standardize chip design. Why couldn’t you program a machine to design circuits, they wondered. “Once you can write a program to do something,” Mead declared, “you don’t need anybody’s tool kit, you write your own.”

康威和米德最终制定了一套数学“设计规则”,为计算机程序自动化芯片设计铺平了道路。使用 Conway 和 Mead 的方法,设计人员不必勾勒出每个晶体管的位置,而是可以从他们的技术使之成为可能的“可互换部件”库中提取。米德喜欢把自己想象成约翰内斯·古腾堡,他的图书生产机械化让作家专注于写作,而印刷商则专注于印刷。康威很快就被麻省理工学院邀请教授这种芯片设计方法的课程。她的每个学生都设计了自己的芯片,然后将设计运送到制造工厂进行制造。六周后,康威的学生从未涉足晶圆厂,他们通过邮件收到了功能齐全的芯片。古腾堡时刻已经到来。

Conway and Mead eventually drew up a set of mathematical “design rules,” paving the way for computer programs to automate chip design. With Conway and Mead’s method, designers didn’t have to sketch out the location of each transistor but could draw from a library of “interchangeable parts” that their technique made possible. Mead liked to think of himself as Johannes Gutenberg, whose mechanization of book production had let writers focus on writing and printers on printing. Conway was soon invited by MIT to teach a course on this chip design methodology. Each of her students designed their own chips, then shipped the design to a fabrication facility for manufacturing. Six weeks later, having never stepped foot in a fab, Conway’s students received fully functioning chips in the mail. The Gutenberg moment had arrived.

没有人比五角大楼对很快被称为“米德康威革命”更感兴趣。DARPA 资助了一项计划,让大学研究人员将芯片设计发送到尖端晶圆厂生产。尽管 DARPA 以资助未来武器系统而闻名,但在半导体方面,DARPA 仍将重点放在建设教育基础设施上,以便美国拥有芯片设计人员供应充足。DARPA 还帮助大学获得先进的计算机,并与行业官员和学者召开研讨会,讨论关于优质葡萄酒的研究问题。帮助DARPA 认为,公司和教授保持摩尔定律对美国的军事优势至关重要。

No one was more interested in what soon became known as the “Mead-Conway Revolution” than the Pentagon. DARPA financed a program to let university researchers send chip designs to be produced at cutting-edge fabs. Despite its reputation for funding futuristic weapons systems, when it came to semiconductors DARPA focused as much on building educational infrastructure so that America had an ample supply of chip designers. DARPA also helped universities acquire advanced computers and convened workshops with industry officials and academics to discuss research problems over fine wine. Helping companies and professors keep Moore’s Law alive, DARPA reasoned, was crucial to America’s military edge.

芯片行业还资助了大学对芯片设计技术的研究,成立了半导体研究公司,向卡内基梅隆大学和加州大学伯克利分校等大学分配研究经费。在 1980 年代,来自这两所大学的一批学生和教师创立了一系列初创公司,这些初创公司创造了一个以前从未存在过的新行业——用于半导体设计的软件工具。今天,每家芯片公司都使用来自三个芯片设计软件公司的工具,这些公司分别是由这些 DARPA 和 SRC 资助项目的校友创立和建造。

The chip industry also funded university research on chip design techniques, establishing the Semiconductor Research Corporation to distribute research grants to universities like Carnegie Mellon and the University of California, Berkeley. Over the 1980s, a cadre of students and faculty from these two universities founded a series of startups that created a new industry—software tools for semiconductor design—that had never previously existed. Today, every chip company uses tools from each of three chip design software companies that were founded and built by alumni of these DARPA- and SRC-funded programs.

DARPA 还支持研究第二组挑战的研究人员:为芯片不断增长的处理能力寻找新用途。无线通信专家 Irwin Jacobs 就是这样的研究人员之一。雅各布斯出生在马萨诸塞州的一个餐馆老板家庭,在爱上电气工程之前,他曾计划跟随父母进入酒店业。他在 1950 年代玩弄真空管和 IBM 计算器。在 MIT 攻读硕士学位期间,Jacobs 研究了天线和电磁理论,并决定将研究重点放在信息论上——研究如何存储和传达信息。

DARPA also backed researchers studying a second set of challenges: finding new uses for chips’ growing processing power. Irwin Jacobs, an expert in wireless communication, was one such researcher. Born in Massachusetts to a family of restaurant owners, Jacobs had planned to follow his parents into the hospitality industry before falling in love with electrical engineering. He spent the 1950s playing around with vacuum tubes and IBM calculators. While pursuing his master’s degree at MIT, Jacobs studied antennas and electromagnetic theory and decided to focus his research on information theory—the study of how information can be stored and communicated.

几十年来,无线电一直在进行无线传输,但对无线通信的需求不断增长,频谱空间有限。如果您想要一个 99.5 FM 的广播电台,您必须确保已经没有 99.7 的广播电台,否则干扰会使您无法理解。同样的原理也适用于其他形式的无线电通信。包含在给定频谱片段中的信息越多,由于混乱的信号从建筑物反弹并在它们通过空域冲向无线电接收器时相互干扰而产生的错误空间就越小。

Radios had been transmitting wirelessly for decades, but the demands for wireless communication were growing and spectrum space was limited. If you wanted a radio station at 99.5 FM, you had to ensure there wasn’t one at 99.7 already, or the interference would make yours incomprehensible. The same principle applied to other forms of radio communication. The more information that was packed into a given slice of spectrum, the less room there was for error created by muddled signals bouncing off buildings and interfering with each other as they careened through airspace toward a radio receiver.

Jacobs 在加州大学圣地亚哥分校的长期同事 Andrew Viterbi 于 1967 年设计了一种复杂的算法来解码一组杂乱无章的数字信号,这些信号在嘈杂的电波中回荡。它是被科学家称赞为优秀的理论,但 Viterbi 的算法在实践中似乎很难使用。普通无线电将拥有运行复杂算法的计算能力的想法似乎不可信。

Jacobs’s longtime University of California, San Diego, colleague Andrew Viterbi had devised a complex algorithm in 1967 to decode a messy set of digital signals reverberating through noisy airwaves. It was praised by scientists as an excellent piece of theory, but Viterbi’s algorithm seemed difficult to use in practice. The idea that normal radios would ever have the computing power to run complicated algorithms seemed implausible.

1971 年,雅各布斯飞往佛罗里达州的圣彼得堡,参加一个研究通信理论的学者会议。许多教授阴郁地得出结论,他们的学术子领域——将数据编码成无线电波——已经达到了它的实际极限。在无法分类和解释之前,无线电频谱只能容纳有限数量的信号。Viterbi 的算法提供了一种将更多数据打包到同一个无线电频谱中的理论方法,但没有人具备大规模应用这些算法的计算能力。通过空中发送数据的过程似乎碰壁了。“编码已经死了,”一位教授宣称。

In 1971, Jacobs flew to St. Petersburg, Florida, to attend a conference of academics working on communications theory. Many of the professors had glumly concluded that their scholarly subfield—encoding data into radio waves—had reached its practical limits. The radio spectrum could only hold a limited number of signals before they became impossible to sort and interpret. Viterbi’s algorithms provided a theoretical way to pack more data into the same radio spectrum, but no one had the computing power to apply these algorithms at scale. The process of sending data through the air seemed to have hit a wall. “Coding is dead,” one professor declared.

雅各布斯完全不同意。他从后排站起身来,高高举起一个小筹码,说道:“这就是未来。” 雅各布斯意识到,芯片的进步如此之快,以至于他们很快就能够在相同的频谱空间中对更多数量级的数据进行编码。由于一平方英寸硅片上的晶体管数量呈指数增长,因此可以通过给定无线电频谱切片发送的数据量也即将起飞。

Jacobs completely disagreed. Standing up from the back row, he held aloft a small chip and declared: “This is the future.” Chips, Jacobs realized, were improving so rapidly that they’d soon be able to encode orders of magnitude more data in the same spectrum space. Because the number of transistors on a square inch of silicon was increasing exponentially, the amount of data that could be sent through a given slice of the radio spectrum was about to take off, too.

Jacobs、Viterbi 和几位同事成立了一家名为 Qualcomm 的无线通信业务——质量通信——他们押注功能越来越强大的微处理器会让他们将更多的信号填充到现有的频谱带宽中。Jacobs 最初从 DARPA 和 NASA 获得了建造太空通信系统的合同。1980 年代后期,高通公司多元化进入民用市场,为货运行业推出了卫星通信系统。但即使到了 1990 年代初,使用芯片通过空中发送大量数据似乎还是一项利基业务。

Jacobs, Viterbi, and several colleagues set up a wireless communications business called Qualcomm—quality communications—betting that ever-more-powerful microprocessors would let them stuff more signals into existing spectrum bandwidth. Jacobs initially won contracts from DARPA and NASA to build space communications systems. In the late 1980s, Qualcomm diversified into the civilian market, launching a satellite communications system for the trucking industry. But even by the early 1990s, using chips to send large quantities of data through the air seemed like a niche business.

对于像欧文·雅各布斯这样的教授出身的企业家来说,DARPA 的资金和国防部的合同对于维持他的创业公司的生存至关重要。但只有一些政府项目奏效了。例如,Sematech 拯救美国光刻机领导者的努力就以惨败告终。政府的努力不是在试图重振失败的公司时才奏效,而是在他们利用美国已有的优势时,提供资金让研究人员将聪明的想法变成原型产品。如果国会议员得知 DARPA(表面上是一个国防机构)在计算机科学教授关于芯片设计的理论时,他们肯定会大发雷霆。但正是这样的努力缩小了晶体管,发现了半导体的新用途,促使新客户购买它们,并为下一代更小的晶体管提供了资金。在半导体设计方面,世界上没有哪个国家拥有更好的创新生态系统。到 1980 年代末,

For a professor-turned-entrepreneur like Irwin Jacobs, DARPA funding and Defense Department contracts were crucial in keeping his startups afloat. But only some government programs worked. Sematech’s effort to save America’s lithography leader was an abject failure, for example. Government efforts were effective not when they tried to resuscitate failing firms, but when they capitalized on pre-existing American strengths, providing funding to let researchers turn smart ideas into prototype products. Members of Congress would no doubt have been furious had they learned that DARPA—ostensibly a defense agency—was wining and dining professors of computer science as they theorized about chip design. But it was efforts like these that shrank transistors, discovered new uses for semiconductors, drove new customers to buy them, and funded the subsequent generation of smaller transistors. When it came to semiconductor design, no country in the world had a better innovation ecosystem. By the end of the 1980s, a chip with a million transistors—unthinkable in the early 1970s, when Lynn Conway had arrived in Silicon Valley—had become a reality, when Intel announced its 486 microprocessor, a small piece of silicon packed with 1.2 million microscopic switches.

第 25 章 克格勃总局 T

CHAPTER 25 The KGB’s Directorate T

V ladimir Vetrov 是一名克格勃间谍,但他的生活更像是契诃夫的故事,而不是詹姆斯邦德的电影。他的克格勃工作很官僚,他的情妇远非超模,他的妻子对她的西施小狗比对他更深情。到 1970 年代末,维特罗夫的职业生涯和他的生活陷入了死胡同。他鄙视自己的办公桌工作,并被老板忽视。他讨厌他的妻子,因为她和他的一个朋友有染。为了消遣,他逃到莫斯科北部一个村庄的小木屋里,那里非常质朴,没有电。或者他只是留在莫斯科喝醉。

Vladimir Vetrov was a KGB spy, but his life felt more like a Chekhov story than a James Bond film. His KGB work was bureaucratic, his mistress far from a supermodel, and his wife more affectionate toward her shih tzu puppies than toward him. By the end of the 1970s, Vetrov’s career, and his life, had hit a dead end. He despised his desk job and was ignored by his bosses. He detested his wife, who was having an affair with one of his friends. For recreation, he escaped to his log cabin in a village north of Moscow, which was so rustic that there was no electricity. Or he’d simply stay in Moscow and get drunk.

维特洛夫的生活并不总是那么乏味。在 1960 年代初期,他在巴黎获得了一个外交职位,作为一名“对外贸易官员”,根据肖金部长的“复制”战略,他的任务是收集法国高科技产业的秘密。1963 年,也就是苏联建立了研究微电子学的科学家之城泽列诺格勒的同一年,克格勃成立了一个新的部门,即 Directorate T,代表teknologia。任务:“获取西方设备和技术,”中央情报局的一份报告警告说,“并提高其生产集成电路的能力。”

Vetrov’s life hadn’t always been so dull. In the early 1960s, he’d earned a plum foreign posting in Paris, where as a “foreign trade official” he was tasked with gathering secrets from France’s high-tech industries, per Minister Shokin’s “copy it” strategy. In 1963, the same year the USSR established Zelenograd, the city of scientists working on microelectronics, the KGB established a new division, Directorate T, which stood for teknologia. The mission: “acquire Western equipment and technology,” a CIA report warned, “and improve its ability to produce integrated circuits.”

据报道,在 1980 年代初期,克格勃雇佣了大约 1000 人来窃取外国技术。大约三百人在外国岗位工作,其余大部分在克格勃八楼将总部设在莫斯科的卢比扬卡广场,坐落在斯大林时代的监狱和酷刑室的顶部。其他苏联情报机构,如军方的 GRU,也有专门从事技术盗窃的间谍。据报道,苏联驻旧金山领事馆有一个由 60 名特工组成的小组,专门针对硅谷的科技公司。他们直接偷了芯片,然后从黑市购买,这些黑市由盗贼提供,比如一个名叫“One Eyed Jack”的人,他于 1982 年在加利福尼亚被捕,并被指控通过将芯片藏在皮夹克中从英特尔工厂窃取芯片。苏联间谍还勒索西方人获得先进技术。至少一名居住在莫斯科的英国计算机公司的英国雇员从高层公寓的窗户“掉下”身亡。

In the early 1980s, the KGB reportedly employed around one thousand people to steal foreign technology. Around three hundred worked at foreign posts, with most of the rest on the eighth floor of the KGB’s imposing headquarters on Moscow’s Lubyanka Square, sitting atop the Stalin-era prison and torture chambers. Other Soviet intelligence services, like the military’s GRU, also had spies who focused on technology theft. The Soviet consulate in San Francisco reportedly had a team of sixty agents targeting the tech firms of Silicon Valley. They stole chips directly and bought them from the black market, supplied by thieves like the man called “One Eyed Jack,” who was caught in California in 1982 and accused of stealing chips from an Intel facility by hiding them in his leather jacket. Soviet spies also blackmailed Westerners with access to advanced technology. At least one British employee of a UK computer company living in Moscow died after “falling” from the window of his high-rise apartment building.

1982 年秋天,一群罗德岛渔民在从北大西洋水域拉出一个奇怪的金属浮标后发现,间谍活动在苏联半导体中继续发挥着重要作用。他们没想到会在运输过程中捡到先进的芯片. 然而,当这个神秘的浮标被送到军事实验室时,它被确定为苏联的监听设备,使用了德州仪器 5400 系列半导体的完美复制品。在英特尔将微处理器商业化之后,同时,肖金部长关闭了一家试图生产类似设备的苏联研究单位,转而复制美国的微处理器。

Spying continued to play a fundamental role in Soviet semiconductors, as a group of Rhode Island fishermen discovered after pulling a strange metallic buoy out of the waters of the North Atlantic in fall 1982. They hadn’t expected to pick up advanced chips in their haul. When the mysterious buoy was sent to a military lab, however, it was identified as a Soviet listening device that used perfect replicas of Texas Instruments Series 5400 semiconductors. After Intel commercialized the microprocessor, meanwhile, Minister Shokin shut down a Soviet research unit trying to produce a similar device, in favor of copying American microprocessors.

然而,“复制”战略远没有苏联监视浮标所暗示的那么成功。窃取英特尔最新芯片的几个例子很容易,甚至将整批集成电路转移到苏联,通常通过中立的奥地利或瑞士的空壳公司。然而,美国的反间谍活动偶尔会揭露苏联在第三国活动的特工,因此这从来都不是可靠的供应来源。

However, the “copy it” strategy was far less successful than Soviet surveillance buoys suggested. It was easy enough to steal a couple examples of Intel’s latest chips, or even to have an entire shipment of integrated circuits diverted to the USSR, usually via shell companies in neutral Austria or Switzerland. However, American counterintelligence occasionally unmasked the USSR’s agents operating in third countries, so this was never a reliable source of supply.

窃取芯片设计只有在苏联能够大规模生产时才有用。这在冷战初期很难做到,但到 1980 年代几乎不可能做到。随着硅谷在硅芯片上塞进更多的晶体管,建造它们变得越来越稳定更难。克格勃认为其盗窃行动为苏联半导体生产商提供了非凡的秘密,但获得新芯片的副本并不能保证苏联工程师能够生产它。克格勃也开始窃取半导体制造设备。中央情报局声称,苏联已经获得了半导体制造过程的几乎所有方面,包括用于制备半导体制造所需材料的 900 台西方机器;八百台光刻机和蚀刻机;三百台机器掺杂、封装和测试芯片。

Stealing chip designs was only useful if they could be produced at scale in the USSR. This was difficult to do during the early Cold War but almost impossible by the 1980s. As Silicon Valley crammed more transistors onto silicon chips, building them became steadily harder. The KGB thought its campaign of theft provided Soviet semiconductor producers with extraordinary secrets, but getting a copy of a new chip didn’t guarantee Soviet engineers could produce it. The KGB began stealing semiconductor manufacturing equipment, too. The CIA claimed that the USSR had acquired nearly every facet of the semiconductor manufacturing process, including nine hundred Western machines for preparing materials needed for semiconductor fabrication; eight hundred machines for lithography and etching; and three hundred machines each for doping, packaging, and testing chips.

然而,一家工厂需要一整套设备,当机器出现故障时,他们需要备件。有时可以在苏联生产外国机器的备件,但这带来了新的低效率和缺陷。盗窃和复制系统的效果一直不足以让苏联军方领导人相信他们拥有稳定的优质芯片供应,因此他们尽量减少在军事系统中使用电子设备和计算机。

However, a factory needed a full suite of equipment, and when machines broke down, they needed spare parts. Sometimes spare parts for foreign machines could be produced in the USSR, but this introduced new inefficiencies and defects. The system of theft and replication never worked well enough to convince Soviet military leaders they had a steady supply of quality chips, so they minimized the use of electronics and computers in military systems.

西方意识到盗窃的规模需要时间。当克格勃于 1965 年首次将维特洛夫派往巴黎时,T 局几乎不为人知。维特罗夫和他的同事从事卧底工作,通常是苏联外贸部的雇员。当苏联特工访问外国研究实验室、结识高管并试图窃取外国工业的秘密时,他们看起来就像是在做外贸官员的“日常工作”。

It took time for the West to realize the scale of the theft. When the KGB first sent Vetrov to Paris in 1965, Directorate T was all but unknown. Vetrov and his colleagues worked undercover, often as employees of the Soviet Ministry of Foreign Trade. When Soviet agents visited foreign research labs, befriended executives, and tried to siphon the secrets of foreign industry, it looked as if they were simply conducting their “day job” as foreign trade officials.

如果 Vetrov 在搬回莫斯科后没有决定为他原本沉闷的生活增添阴谋,T 局的行动可能仍然是国家机密。到了 1980 年代初,他的事业停滞不前,他的婚姻破裂,他的生活也分崩离析。他是像詹姆斯邦德一样的间谍,但更多的办公桌工作和更少的马提尼酒。他决定寄一张明信片给他认识的巴黎熟人,让生活变得更有趣。与法国情报部门有联系。

The operations of Directorate T might have remained a state secret had Vetrov not decided to add intrigue to his otherwise dull existence upon moving back to Moscow. By the early 1980s, his career had stalled, his marriage was ruined, and his life was falling apart. He was a spy like James Bond, but with more desk work and fewer martinis. He decided to make life more interesting by sending a postcard to a Parisian acquaintance who, he knew, was connected with the French intelligence services.

很快,维特罗夫就将数十份关于 T 局的文件交给了他在莫斯科的法国经理。法国情报部门代号他“告别。” 总的来说,他似乎提供了数千页来自克格勃核心的文件,揭露了一个专注于窃取西方工业机密的庞大官僚机构。一个关键的优先事项:“先进的微处理器”,苏联不仅缺乏熟练的工程师,而且缺乏设计尖端处理器所需的软件和生产它们所需的设备。西方间谍惊呆了苏联偷了多少。

Soon Vetrov was passing dozens of documents about Directorate T to his French handler in Moscow. French intelligence code-named him “Farewell.” In total, he appears to have provided thousands of pages of documents from the heart of the KGB, unveiling a vast bureaucracy focused on stealing Western industrial secrets. A key priority: “advanced microprocessors,” for which the Soviet Union lacked not only skilled engineers but also the software needed to design cutting-edge processors and the equipment needed to produce them. Western spies were shocked at just how much the Soviets stole.

在他与法国特工会合的例行公事中,维特罗夫找到了一项新的活动,但他没有找到满足感。法国人从国外送他礼物,让维特罗夫的情妇开心,但维特罗夫真正想要的是让他的妻子爱他。他越来越妄想。1982 年 2 月 22 日,在告诉儿子他打算与情妇断绝关系后,维特罗夫在莫斯科环城公路上停车时,在他的车里反复刺伤了她。直到他被警方逮捕后,克格勃才意识到维特罗夫背叛了他的国家,并将 T 局的秘密交给了西方情报部门。

In his routine of rendezvousing with French agents Vetrov had found a new activity, but he hadn’t found fulfillment. The French provided him with gifts from abroad, to keep Vetrov’s mistress happy, yet what Vetrov really wanted was for his wife to love him. He grew ever more delusional. On February 22, 1982, having told his son he planned to break off the relationship with his mistress, Vetrov stabbed her repeatedly in his car while parked along Moscow’s ring road. Only after he was apprehended by police did the KGB realize Vetrov had betrayed his country and handed the secrets of Directorate T to Western intelligence.

法国人迅速与美国和其他盟国情报机构分享了有关维特罗夫的信息。作为回应,里根政府启动了“出埃及记”行动,加强了海关对先进技术的检查。到 1985 年,该计划已查获价值约 6 亿美元的货物,并导致约 1000 人被捕。然而,在半导体方面,里根政府声称已经阻止了“美国技术向苏联大量流失”的说法可能夸大了更严格控制的影响。苏联的“复制”战略实际上使美国受益,保证了苏联面临持续的技术落后。1985 年,中央情报局对苏联的微处理器进行了研究,发现苏联生产了类似发条的英特尔和摩托罗拉芯片的复制品。他们是总是落后五年。

The French quickly shared information about Vetrov with U.S. and other allied intelligence services. The Reagan administration responded by launching Operation Exodus, which tightened customs checks on advanced technology. By 1985, the program had seized around $600 million worth of goods and resulted in around one thousand arrests. However, when it came to semiconductors, the Reagan administration’s claim to have stopped the “massive hemorrhage of American technology to the Soviet Union” probably overstated the impact of tighter controls. The USSR’s “copy it” strategy had actually benefitted the United States, guaranteeing the Soviets faced a continued technological lag. In 1985, the CIA conducted a study of Soviet microprocessors and found that the USSR produced replicas of Intel and Motorola chips like clockwork. They were always half a decade behind.

第 26 章 “大规模杀伤性武器”:抵消的影响

CHAPTER 26 “Weapons of Mass Destruction” : The Impact of the Offset

远程、高精度、末端制导的战斗系统、无人驾驶飞行器和全新的电子控制系统,”苏联元帅尼古拉·奥加科夫预测,将把常规炸药转变为“大规模杀伤性武器”。奥加科夫于1977年至1984年担任苏联军队总参谋长。在西方,他最出名的是在1983年苏联意外击落一架来自韩国的民用客机后,领导媒体攻势。与其认错,他指责这架飞机的飞行员正在执行“经过深思熟虑的、经过周密计划的情报任务”,并宣称这架客机是“要它。” 这不是一个可能让奥加科夫在西方赢得任何朋友的信息,但这对他来说可能没什么影响,因为他的人生目标是准备与美国开战。

“Long-range, highly accurate, terminally guided combat systems, unmanned flying machines, and qualitatively new electronic control systems,” Soviet Marshal Nikolai Ogarkov predicted, would transform conventional explosives into “weapons of mass destruction.” Ogarkov served as chief of the general staff of the Soviet military from 1977 to 1984. In the West, he was most famous for leading the media offensive after the Soviets accidentally shot down a civilian airliner from South Korea in 1983. Rather than admit a mistake, he accused the plane’s pilots of being on a “deliberate, thoroughly planned intelligence mission” and declared that the airliner was “asking for it.” This wasn’t a message likely to win Ogarkov any friends in the West, but that was likely of little consequence to him since his life purpose was preparing for war with the United States.

苏联在开发冷战初期的关键技术、制造强大的火箭和强大的核武库方面一直在跟上美国人的步伐。现在,肌肉正在被计算机化的大脑所取代。当谈到硅支撑这种新的军事力量动力的筹码,苏联已经无可救药地落后了。1980 年代的一个流行的苏联笑话讲述了一位克里姆林宫官员自豪地宣称:“同志,我们制造了世界上最大的微处理器!”

The Soviet Union had kept up with the Americans in the race to develop the crucial technologies of the early Cold War, building powerful rockets and a formidable nuclear stockpile. Now brawn was being replaced by computerized brains. When it came to the silicon chips undergirding this new driver of military power, the Soviet Union had fallen hopelessly behind. One popular Soviet joke from the 1980s recounted a Kremlin official who declared proudly, “Comrade, we have built the world’s biggest microprocessor!”

根据坦克或部队数量等传统指标,苏联在 1980 年代初期具有明显优势。奥加科夫的看​​法不同:质量超过数量。他专注于美国精确武器构成的威胁。结合更好的监视和通信工具,准确打击数百甚至数千英里外目标的能力正在产生“军事技术革命,”奥加科夫对任何愿意倾听的人争辩道。真空管制导的“麻雀”导弹在越南上空丢失 90% 的目标的日子早已一去不复返了。苏联的坦克比美国多得多,但奥加科夫意识到他的坦克很快就会在与美国的战斗中变得脆弱许多倍

By traditional metrics like numbers of tanks or troops, the Soviet Union had a clear advantage in the early 1980s. Ogarkov saw things differently: quality was overtaking quantity. He was fixated on the threat posed by America’s precision weapons. Combined with better surveillance and communication tools, the ability to strike targets accurately hundreds or even thousands of miles away was producing a “military-technical revolution,” Ogarkov argued to anyone who’d listen. The days of vacuum tube−guided Sparrow missiles missing 90 percent of their targets in the skies over Vietnam were long gone. The Soviet Union had many more tanks than the United States, but Ogarkov realized his tanks would soon be many times more vulnerable in a fight with the U.S.

账单佩里的“抵消战略”奏效了,苏联没有回应。它缺乏美国和日本芯片制造商生产的小型化电子设备和计算能力。Zelenograd 和其他苏联芯片制造设施无法跟上。尽管佩里推动五角大楼接受摩尔定律,但苏联芯片制造的不足之处却教会了该国的武器设计者尽可能限制复杂电子设备的使用。这在 1960 年代是一种可行的方法,但到了 1980 年代,这种不愿跟上微电子技术进步的态度保证了即使美国武器正在学习思考,苏联系统仍将保持“愚蠢”。美国在 1960 年代初将由德州仪器 (TI) 芯片驱动的制导计算机安装在民兵 II 导弹上,但苏联第一台使用集成电路的导弹制导计算机直到 1971 年才进行了测试。

Bill Perry’s “offset strategy” was working, and the Soviet Union didn’t have a response. It lacked the miniaturized electronics and computing power that American and Japanese chipmakers produced. Zelenograd and other Soviet chipmaking facilities couldn’t keep up. Whereas Perry pushed the Pentagon to embrace Moore’s Law, the inadequacies of Soviet chipmaking taught the country’s weapons designers to limit use of complex electronics whenever possible. This was a viable approach in the 1960s, but by the 1980s this unwillingness to keep pace with advances in microelectronics guaranteed Soviet systems would remain “dumb” even as American weapons were learning to think. The U.S. had put a guidance computer powered by Texas Instruments’ chips onboard the Minuteman II missile in the early 1960s, but the Soviets’ first missile guidance computer using integrated circuits wasn’t tested until 1971.

习惯于低质量的微电子技术的苏联导弹设计师设计了精心设计的解决方法。甚至他们插入导航计算机的数学也更简单,以最大限度地减少车载计算机的压力。苏联弹道导弹通常被告知沿着特定的飞行路径飞向他们的目标,如果导弹偏离,制导计算机会调整导弹使其回到预编程的路线上。相比之下,到 1980 年代,美国的导弹计算出自己到达目标的路径。

Accustomed to low-quality microelectronics, Soviet missile designers devised elaborate workarounds. Even the mathematics they plugged into their guidance computers was simpler, to minimize the strain on the onboard computer. Soviet ballistic missiles were generally told to follow a specific flight path toward their target, with the guidance computer adjusting the missile to put it back on the preprogrammed route if it deviated. By contrast, by the 1980s, American missiles calculated their own path to the target.

到 1980 年代中期,据公开估计,美国的新型 MX 导弹有 50% 的时间在其目标 364 英尺范围内着陆。根据一位前苏联国防官员的估计,一种大致可比的苏联导弹 SS-25 平均落在其目标的 1200 英尺范围内。在冷战军事规划者的严酷逻辑中,几百英尺的差异非常重要。摧毁一座城市很容易,但两个超级大国都希望有能力摧毁对方的核武库。即使是核弹头也需要合理的直接打击才能使硬化的导弹发射井失效。足够的直接打击,一方可能会在出其不意的第一次打击中危及对手的核力量。苏联最悲观的估计表明,如果美国在 1980 年代发动第一次核打击,摧毁了 98% 的苏联洲际弹道导弹。

By the mid-1980s, America’s new MX missile was publicly estimated to land within 364 feet of its target 50 percent of the time. A roughly comparable Soviet missile, the SS-25, on average fell within twelve hundred feet of its target, according to estimates from a former Soviet defense official. In the grim logic of Cold War military planners, a difference of several hundred feet mattered hugely. It was easy enough to destroy a city, but both superpowers wanted the ability to knock out each other’s nuclear arsenals. Even nuclear warheads needed a reasonably direct hit to disable a hardened missile silo. Enough direct hits, and one side could potentially compromise the adversary’s nuclear forces in a surprise first strike. The most pessimistic Soviet estimates suggested that if the U.S. launched a nuclear first strike in the 1980s, it could have disabled or destroyed 98 percent of Soviet ICBMs.

苏联没有任何错误的余地。苏联军方还有两个可以对美国发动核攻击的系统:远程轰炸机和导弹潜艇。轰炸机舰队被广泛认为是最弱的运载系统,因为它们可以在起飞后不久被雷达识别并在发射核武器之前被击落。相比之下,美国的核导弹潜艇实际上是探测不到的,因此是无敌的。苏联潜艇的安全性较低,因为美国正在学习应用计算能力来提高其潜艇探测系统的准确性。

The USSR didn’t have any margin for error. The Soviet military had two other systems that could launch a nuclear attack on America: long-range bombers and missile submarines. Bomber fleets were widely agreed to be the weakest delivery system because they could be identified by radar shortly after taking off and shot down before launching their nuclear weapons. America’s nuclear missile submarines, by contrast, were practically undetectable and therefore invincible. Soviet submarines were less secure, because the U.S. was learning to apply computing power to make its submarine detection systems far more accurate.

寻找潜艇的挑战是理解声波的杂音。声音以不同的角度从海底反弹,并根据温度或鱼群的存在而在水中以不同的方式折射。到 1980 年代初,公开承认美国已将其潜艇传感器插入 Illiac IV,这是最强大的超级计算机之一,也是第一个使用半导体存储芯片,由 Fairchild 制造。Illiac IV 和其他处理中心通过卫星连接到船舶、飞机和直升机上的一系列传感器,以跟踪苏联潜艇,这些潜艇极易受到美国侦察。

The challenge in finding a submarine is to make sense of a cacophony of sound waves. Sound bounces off the seafloor at different angles and refracts differently through water depending on the temperature or the presence of schools of fish. By the early 1980s it was publicly admitted that the U.S. had plugged its submarine sensors into the Illiac IV, one of the most powerful supercomputers and the first using semiconductor memory chips, which were built by Fairchild. Illiac IV and other processing centers were connected via satellite to an array of sensors on ships, planes, and helicopters to track Soviet subs, which were highly vulnerable to American detection.

当奥加科夫计算这些数字时,他得出的结论是,美国在导弹精度、反潜战、监视和指挥控制方面的半导体动力优势可以使突然袭击威胁到苏联核武库的生存能力。核武器应该是最终的保险政策,但苏联军方现在觉得正如一位将军所说,“战略武器明显逊色”。

When Ogarkov ran the numbers, he concluded that America’s semiconductor-powered advantage in missile accuracy, antisubmarine warfare, surveillance, and command and control could enable a surprise strike to threaten the survivability of the Soviet nuclear arsenal. Nukes were supposed to be the ultimate insurance policy, but the Soviet military now felt “substantially inferior in strategic weapons,” as one general put it.

苏联军事领导人也担心一场常规战争。军事分析家此前认为,苏联在坦克和部队数量方面的优势在常规战争中提供了决定性优势。然而,首先在越南上空使用的宝石路炸弹得到了一套新的制导系统的补充。战斧巡航导弹可以深入苏联领土。苏联国防规划者担心美国常规武装巡航导弹和隐形轰炸机可能会削弱苏联对其核力量的指挥和控制。挑战受到威胁苏维埃国家的生存。

Soviet military leaders feared a conventional war, too. Military analysts previously thought the Soviets’ superiority in numbers of tanks and troops provided a decisive advantage in a conventional war. However, the Paveway bomb first used over Vietnam had been supplemented by a suite of new guided systems. Tomahawk cruise missiles could strike deep into Soviet territory. Soviet defense planners feared American conventionally armed cruise missiles and stealth bombers could disable Soviet command and control over their nuclear forces. The challenge threatened the very survival of the Soviet state.

克里姆林宫想振兴其微电子产业,但不知道如何去做。1987年,苏联领导人米哈伊尔·戈尔巴乔夫访问泽列诺格勒并呼吁在城市的工作中“更纪律”。纪律是硅谷成功的一部分,查理·斯波克对生产力的执着和安迪·格罗夫的偏执很明显。然而,单靠纪律并不能解决苏联人的基本问题。

The Kremlin wanted to revitalize its microelectronics industry but didn’t know how to do so. In 1987, Soviet leader Mikhail Gorbachev visited Zelenograd and called for “more discipline” in the city’s work. Discipline was part of Silicon Valley’s success, evident in Charlie Sporck’s fixation on productivity and Andy Grove’s paranoia. However, discipline alone couldn’t solve the Soviets’ basic problems.

一个问题是政治干预。1980 年代后期,尤里奥索金在里加半导体厂被免职。克格勃要求他解雇几名雇员,其中一名曾给捷克斯洛伐克的一名妇女寄信,另一名拒绝为克格勃提供线人,另一名是犹太人。当奥索金拒绝惩罚这些工人的“罪行”时,克格勃将他赶下台并试图让他的妻子也被解雇。平时设计芯片已经够难的了。在与克格勃作战时这样做是不可能的。

One issue was political meddling. In the late 1980s, Yuri Osokin was removed from his job at the Riga semiconductor plant. The KGB had demanded that he fire several of his employees, one of whom had mailed letters to a woman in Czechoslovakia, a second who refused to work as an informant for the KGB, and a third who was a Jew. When Osokin refused to punish these workers for their “crimes,” the KGB ousted him and tried to have his wife fired, too. It was hard enough to design chips in normal times. Doing so while battling the KGB was impossible.

第二个问题是过度依赖军事客户。美国、欧洲和日本的消费市场蓬勃发展,推动了芯片需求。民用半导体市场为半导体供应链的专业化提供了资金支持,创建了在从超纯硅晶片到光刻设备中的先进光学等各个方面都具有专业知识的公司。苏联几乎没有消费市场,所以它只生产了西方制造的芯片的一小部分。一位苏联消息人士估计,仅日本就花费了微电子领域的资本投资是苏联的八倍。

A second issue was overreliance on military customers. The U.S., Europe, and Japan had booming consumer markets that drove chip demand. Civilian semiconductor markets helped fund the specialization of the semiconductor supply chain, creating companies with expertise in everything from ultra-pure silicon wafers to the advanced optics in lithography equipment. The Soviet Union barely had a consumer market, so it produced only a fraction of the chips built in the West. One Soviet source estimated that Japan alone spent eight times as much on capital investment in microelectronics as the USSR.

最后的挑战是苏联人缺乏国际供应链。硅谷与美国的冷战盟友合作,打造了超高效的全球化分工。日本主导了存储芯片的生产,美国生产了更多的微处理器,而日本的尼康和佳能以及荷兰的ASML瓜分了光刻设备市场。东南亚的工人进行了最后组装的大部分工作。美国、日本和欧洲的公司在这种分工中争夺自己的地位,但他们都受益于将研发成本分散到比苏联更大的半导体市场的能力。

A final challenge was that the Soviets lacked an international supply chain. Working with America’s Cold War allies, Silicon Valley had forged an ultra-efficient globalized division of labor. Japan led the production of memory chips, the U.S. produced more microprocessors, while Japan’s Nikon and Canon and the Netherland’s ASML split the market for lithography equipment. Workers in Southeast Asia conducted much of the final assembly. American, Japanese, and European companies jostled over their position in this division of labor, but they all benefitted from the ability to spread R&D costs over a far larger semiconductor market than the USSR ever had.

苏联只有少数几个盟友,其中大多数没有多大帮助。苏联主导的东德拥有与泽列诺格勒一样先进的芯片工业,在 1980 年代中期,利用精密制造的悠久传统以及由位于耶拿市的卡尔蔡司公司。东德的芯片产量在 1980 年代后期迅速增长,但该行业只能生产不如日本先进的存储芯片,价格是日本的十倍。先进的西方制造设备仍然难以获得,而东德没有硅谷公司在亚洲雇佣的廉价劳动力。

The USSR had only a handful of allies, most of whom weren’t much help. Soviet-dominated East Germany, which had a chip industry as advanced as Zelenograd, made a last-ditch effort in the mid-1980s to revitalize its semiconductor sector, drawing on a long tradition of precision manufacturing as well as world-leading optics produced by the Carl Zeiss company in the city of Jena. East German chip output grew rapidly in the late 1980s, but the industry was only able to produce memory chips less advanced than Japan’s, at ten times the price. Advanced Western manufacturing equipment remained hard to access, while East Germany had none of the cheap labor that Silicon Valley firms hired across Asia.

苏联重振其芯片制造商的努力完全失败了。尽管进行了大规模的间谍活动,并且向泽列诺格勒这样的研究机构投入了大量资金,但苏联人和他们的社会主义盟友都无法赶上。就在克里姆林宫对比尔·佩里“抵消”的回应开始浮出水面时,世界对波斯湾战场上战争的未来有了可怕的一瞥。

The Soviet Union’s effort to reinvigorate its chipmakers failed completely. Neither the Soviets nor their socialist allies could ever catch up, despite vast espionage campaigns and huge sums poured into research facilities like those in Zelenograd. And just as the Kremlin’s response to Bill Perry’s “offset” was beginning to sputter out, the world was given a terrifying glimpse of the future of war on the battlefields of the Persian Gulf.

第 27 章 战争英雄

CHAPTER 27 War Hero

1991 年 1 月 17 日凌晨,第一批美国F -117 隐形轰炸机从沙特阿拉伯的空军基地起飞,黑色的机身迅速消失在黑暗的沙漠天空中。他们的目标:巴格达。美国自越战以来就没有打过一场大战,但现在它在沙特阿拉伯北部边境拥有数十万军队,数万辆坦克等待命令向前冲锋,数十艘海军舰艇驻扎在海上,他们的枪支和导弹电池瞄准伊拉克。领导进攻的美国将军,Norman Schwarzkopf 是一名训练有素的步兵,曾在越南服役过两次。这一次,他信赖防区外武器来发动第一击。

Early in the morning on January 17, 1991, the first wave of American F-117 stealth bombers took off from their airbases in Saudi Arabia, their black airframes quickly disappearing in the dark desert sky. Their target: Baghdad. The United States hadn’t fought a major war since Vietnam, but now it had several hundred thousand troops along Saudi Arabia’s northern border, tens of thousands of tanks awaiting orders to storm forward, dozens of naval ships positioned offshore, their guns and missile batteries aimed at Iraq. The American general leading the assault, Norman Schwarzkopf, was an infantryman by training, having served two tours in Vietnam. This time, he was trusting in stand-off weapons to deliver the first strike.

位于巴格达拉希德街的 12 层高的电话交换大楼是唯一被认为足以被两架 F-117 攻击的重要目标。施瓦茨科普夫将军的战争计划依赖于它被摧毁,摧毁伊拉克的部分通信基础设施。两架飞机瞄准他们的目标,释放了 2000 磅重的宝石路激光制导炸弹,炸穿了设施并将其点燃。突然,CNN 驻巴格达记者的电视画面变暗了。施华蔻的飞行员受到了打击。几乎同时,近海海军舰艇发射的 116 枚战斧巡航导弹击中了巴格达及其周边地区的目标。波斯湾战争已经开始。

The twelve-story tall telephone exchange building on Baghdad’s Rashid Street was the only target deemed important enough to be attacked by two F-117s. General Schwarzkopf’s war plan depended on it being destroyed, knocking out part of Iraq’s communications infrastructure. The two planes homed in on their target, releasing two-thousand-pound Paveway laser-guided bombs that tore through the facility and set it aflame. Suddenly the TV feed of CNN’s reporters in Baghdad went dark. Schwarzkopf’s pilots had scored a hit. Almost simultaneously, 116 Tomahawk cruise missiles fired from naval ships offshore slammed into their targets in and around Baghdad. The Persian Gulf War had begun.

通讯塔、军事指挥所、空军总部、发电站和萨达姆侯赛因的乡间别墅——第一个美国的空袭试图斩首伊拉克领导人并切断他们的通讯,限制他们追踪战争或与他们的部队沟通的能力。很快他们的军队就陷入了无组织的撤退。美国有线电视新闻网播放了数百枚炸弹和导弹袭击伊拉克坦克的视频。战争看起来像一个电子游戏。但从德克萨斯州观看,韦尔登·沃德知道这种未来技术实际上可以追溯到越南战争。

A communications tower, a military command post, air force headquarters, power stations, and Saddam Hussein’s country retreat—the first U.S. airstrikes sought to decapitate the Iraqi leadership and cut their communications, limiting their ability to track the war or communicate with their forces. Soon their military was in a disorganized retreat. CNN broadcast videos of hundreds of bombs and missiles striking Iraqi tanks. Warfare looked like a video game. But watching from Texas, Weldon Word knew this futuristic technology actually dated to the Vietnam War.

撞击巴格达电话交换机的宝石路激光制导炸弹使用的基本系统设计与 1972 年摧毁清化大桥的第一代宝石路炸弹相同。这些炸弹由少数晶体管、一个激光传感器和几个翅膀绑在一个旧的“哑”炸弹上。到 1991 年,德州仪器已多次更新 Paveway,每个新版本都用更先进的电子设备替换现有电路,减少组件数量,提高可靠性并添加新功能。到波斯湾战争开始时,Paveway 已成为军队的首选武器,原因与英特尔的微处理器在整个计算机行业被广泛使用的原因相同:它们广为人知、易于使用且具有成本效益。铺路石总是很便宜,但它们在 1970 年代和 1980 年代变得更便宜。由于成本低廉,每位飞行员都在训练演习中放弃了宝石路。它们也用途广泛。目标不需要提前选择,可以在战场上选择。与此同时,点击率几乎和电视上看到的一样好。战后进行的空军研究发现,非精确弹药的精确度远低于飞行员通常声称的准确度,而像宝石路炸弹这样的精确弹药实际上比声称的要好。使用激光制导进行炸弹袭击的飞机被击中 几乎和他们在电视上看到的一样好。战后进行的空军研究发现,非精确弹药的精确度远低于飞行员通常声称的准确度,而像宝石路炸弹这样的精确弹药实际上比声称的要好。使用激光制导进行炸弹袭击的飞机被击中 几乎和他们在电视上看到的一样好。战后进行的空军研究发现,非精确弹药的精确度远低于飞行员通常声称的准确度,而像宝石路炸弹这样的精确弹药实际上比声称的要好。使用激光制导进行炸弹袭击的飞机被击中目标是没有制导弹药的同类飞机的 13 倍。

The Paveway laser-guided bombs that slammed into Baghdad’s telephone exchange used the same basic system design as the first generation of Paveways that destroyed the Thanh Hoa Bridge in 1972. Those were built with a handful of transistors, a laser sensor, and a couple of wings strapped to an old “dumb” bomb. By 1991, Texas Instruments had updated the Paveway multiple times, with each new version replacing existing circuitry with more advanced electronics, reducing the number of components, increasing reliability, and adding new features. By the start of the Persian Gulf War, the Paveway had become the military’s weapon of choice for the same reason Intel’s microprocessors were used across the computer industry: they were widely understood, easy to use, and cost-effective. Paveways were always cheap, but they got cheaper over the course of the 1970s and 1980s. Thanks to their low cost, every pilot had dropped Paveways in training exercises. And they were highly versatile, too. Targets didn’t need to be selected in advance but could be chosen on the battlefield. The hit rates, meanwhile, were almost as good as they looked on TV. Air Force studies conducted after the war found that non-precision munitions were far less accurate than pilots often claimed, while precision munitions like the Paveway bombs actually did better than claimed. Planes using laser guidance for their bomb strikes hit thirteen times as many targets as comparable planes without guided munitions.

美国的空中力量在波斯湾战争中被证明是决定性的,它摧毁了伊拉克军队,同时最大限度地减少了美国的伤亡。Weldon Word 收到因发明宝石路、改进其电子设备并降低其成本而获得的奖项,这样每一个都不会像他最初承诺的那样比一辆老爷车贵。美国军方以外的人花了几十年的时间才意识到宝石路和其他类似武器是如何改变战争的。但使用这些炸弹的飞行员知道它们的变革性。“大约有一万个美国人没有因为你们而被杀,”一名空军军官在五角大楼颁奖典礼上告诉 Word。先进的微电子技术和一组绑在炸弹上的机翼已经改变了军事力量的本质。

U.S. airpower proved decisive in the Persian Gulf War, decimating Iraqi forces while minimizing U.S. casualties. Weldon Word received an award for inventing the Paveway, improving its electronics, and driving down its cost so that each one was never more expensive than a jalopy, just as he had originally promised. It took several decades for people outside the U.S. military to realize how the Paveway and other weapons like it were changing war. But pilots who used these bombs knew just how transformative they were. “There are about ten thousand Americans who didn’t get killed because of you guys,” an Air Force officer told Word at the Pentagon award ceremony. Advanced microelectronics and a set of wings strapped to a bomb had transformed the nature of military power.

比尔·佩里 (Bill Perry) 目睹了波斯湾战争的展开,他知道激光制导炸弹只是数十种由集成电路彻底改变的军事系统之一,这些系统可以实现更好的监视、通信和计算能力。波斯湾战争是对佩里“抵消战略”的第一次重大考验,该战略是在越南战争之后设计的,但从未在大规模战斗中部署。

As Bill Perry watched the Persian Gulf War unfold, he knew laser-guided bombs were just one of dozens of military systems that had been revolutionized by integrated circuits, enabling better surveillance, communication, and computing power. The Persian Gulf War was the first major test of Perry’s “offset strategy,” which had been devised after the Vietnam War but never deployed in a sizeable battle.

越战之后的几年里,美军一直在谈论它的新能力,但很多人并没有把它们当回事。在越南指挥美军的威廉威斯特摩兰将军等军事领导人承诺未来的战场将实现自动化。但是,尽管美国对北越拥有广泛的技术优势,但越南战争却是灾难性的。那么,为什么更多的计算能力会改变事情呢?1980 年代,美国军队大部分时间都呆在军营里,除了一些针对利比亚和格林纳达等三流对手的小规模行动。没有人知道五角大楼的先进设备在真实战场上的表现如何。

In the years after Vietnam, the U.S. military had talked about its new capabilities, but many people didn’t take them seriously. Military leaders like General William Westmoreland, who commanded American forces in Vietnam, promised that future battlefields would be automated. But the Vietnam War had gone disastrously despite America’s wide technological advantage over the North Vietnamese. So why would more computing power change things? America’s military mostly sat in its barracks during the 1980s, except for a few small operations against third-rate opponents like Libya and Grenada. No one was sure how the Pentagon’s advanced gadgets would perform on real battlefields.

伊拉克建筑物、坦克和机场被精确武器摧毁的视频让人无法否认:战争的性质正在发生变化。即使是在越南上空错过了大部分目标的真空管动力“响尾蛇”空对空导弹,现在也升级了更强大的基于半导体的制导系统。它们在波斯湾战争中的准确度是越南战争中的六倍。

Videos of Iraqi buildings, tanks, and airfields being destroyed by precision weapons made it impossible to deny: the character of war was changing. Even the vacuum tube−powered Sidewinder air-to-air missiles that had missed most of their targets above Vietnam were now upgraded with more powerful, semiconductor-based guidance systems. They were six times as accurate in the Persian Gulf War as in Vietnam.

佩里在 1970 年代后期推动五角大楼发展的新技术的表现甚至超出了他的预期。伊拉克军队——装备了苏联国防工业生产的一些最好的装备——在美国的进攻面前束手无策。“高科技作品,”佩里宣称。一位军事分析家向媒体解释说:“使这一切奏效的是基于信息而不是火力数量的武器。” “这是硅对钢的胜利,”《纽约时报》的标题宣称。“计算机芯片的战争英雄状态可能,”另一位读者读到。

The new technologies Perry had pushed the Pentagon to develop during the late 1970s performed even beyond his expectations. The Iraqi military—armed with some of the best equipment the Soviet Union’s defense industry produced—was helpless in the face of the American assault. “High-tech works,” Perry proclaimed. “What’s making all this work is weapons based on information instead of the volume of fire power,” one military analyst explained to the media. “It’s the triumph of silicon over steel,” declared a New York Times headline. “War Hero Status Possible for the Computer Chip,” read another.

宝石路炸弹和战斧导弹爆炸的回响在莫斯科和巴格达一样强烈。一位苏联军事分析家宣称,这场战争是一场“技术行动”。另一位说,这是“一场关于电波的斗争”。结果-伊拉克轻而易举的失败——正是奥加科夫所预测的。苏联国防部长德米特里·亚佐夫承认,海湾战争让苏联对其防空能力感到紧张。谢尔盖·阿赫罗梅耶夫元帅在他对一场旷日持久的冲突的预测被伊拉克迅速投降迅速推翻后感到尴尬。美国有线电视新闻网关于美国炸弹在天空中引导自己并撞击伊拉克建筑物墙壁的视频证明了奥加科夫对战争未来的预测。

The reverberations from the explosions of Paveway bombs and Tomahawk missiles were felt as powerfully in Moscow as in Baghdad. The war was a “technological operation,” one Soviet military analyst declared. It was “a struggle over the airwaves,” another said. The result—Iraq’s easy defeat—was exactly what Ogarkov had predicted. Soviet Defense Minister Dmitri Yazov admitted the Gulf War made the Soviet Union nervous about its air defense capabilities. Marshal Sergey Akhromeyev was embarrassed after his predictions of a protracted conflict were promptly disproven by Iraq’s speedy surrender. CNN videos of American bombs guiding themselves through the sky and slamming through the walls of Iraqi buildings proved Ogarkov’s forecasts about the future of war.

第 28 章 “冷战结束了,你赢了”

CHAPTER 28 “The Cold War Is Over and You Have Won”

索尼的 Akio Morita 曾在 1980 年代乘坐飞机环游世界,与亨利·基辛格共进晚餐,在奥古斯塔打高尔夫球,与三边委员会等团体中的其他全球精英交往。他在全球舞台上被视为商业神谕和日本——世界经济崛起大国的代表。森田发现“日本第一”很容易相信,因为他亲身生活在其中。多亏了索尼的随身听和其他消费电子产品,日本变得繁荣,森田变得富有。

Sony’s Akio Morita had spent the 1980s jetting around the world, dining with Henry Kissinger, golfing at Augusta National, hobnobbing with other global elites in groups like the Trilateral Commission. He was treated as a business oracle and a representative of Japan—the world’s rising economic power—on the global stage. Morita found “Japan as Number One” easy to believe in because he was personally living it. Thanks to Sony’s Walkman and other consumer electronics, Japan had become prosperous and Morita had gotten rich.

然后在 1990 年危机袭来。日本金融市场崩盘。经济陷入深度衰退。很快,东京股市的交易价格跌至 1990 年水平的一半。东京的房地产价格进一步下跌。日本的经济奇迹似乎戛然而止。与此同时,美国在商业和战争中都在复苏。短短几年,“日本第一”似乎不再那么准确。日本萎靡不振的案例研究一直被认为是日本工业实力的典范:半导体。

Then in 1990 crisis hit. Japan’s financial markets crashed. The economy slumped into a deep recession. Soon the Tokyo stock market was trading at half its 1990 level. Real estate prices in Tokyo fell even further. Japan’s economic miracle seemed to screech to a halt. Meanwhile, America was resurgent, in business and in war. In just a few short years, “Japan as Number One” no longer seemed very accurate. The case study in Japan’s malaise was the industry that had been held up as exemplary of Japan’s industrial prowess: semiconductors.

现年 69 岁的森田看着日本的财富随着索尼股价的下跌而一落千丈。他知道他的国家的问题比其金融市场更严重。森田在过去的十年向美国人讲授他们需要提高生产质量,而不是专注于金融市场的“金钱游戏”。但随着日本股市崩盘,该国自吹自擂的长期思维不再那么有远见。日本表面上的主导地位是建立在一个不可持续的基础之上的政府支持的过度投资。廉价资本为新半导体工厂的建设提供了资金支持,但也鼓励芯片制造商少考虑利润,多考虑产量。日本最大的半导体公司将 DRAM 产量翻了一番,尽管美光和韩国等成本较低的生产商三星削弱了日本竞争对手。

Morita, now sixty-nine years old, watched Japan’s fortunes decline alongside Sony’s slumping stock price. He knew his country’s problems cut deeper than its financial markets. Morita had spent the previous decade lecturing Americans about their need to improve production quality, not focus on “money games” in financial markets. But as Japan’s stock market crashed, the country’s vaunted long-term thinking no longer looked so visionary. Japan’s seeming dominance had been built on an unsustainable foundation of government-backed overinvestment. Cheap capital had underwritten the construction of new semiconductor fabs, but also encouraged chipmakers to think less about profit and more about output. Japan’s biggest semiconductor firms doubled down on DRAM production even as lower cost producers like Micron and South Korea’s Samsung undercut Japanese rivals.

日本自己的媒体认为半导体行业投资过度,报纸头条警告“鲁莽的投资竞争”和“他们无法停止的投资”。日本存储芯片生产商的 CEO 们不能让自己停止建造新的芯片工厂,即使它们没有盈利。一位日立高管承认,“如果你开始担心”过度投资,“你晚上睡不着。” 只要银行继续放贷,首席执行官们就更容易继续支出,而不是承认他们没有盈利之路。在 1980 年代,美国的公平资本市场并没有被认为是一种优势,但失去融资的风险帮助美国公司保持警觉。日本 DRAM 制造商将从 Andy Grove 的偏执狂或 Jack Simplot 对商品市场波动的智慧中受益。相反,他们都将投资投入同一个市场,保证很少有人能赚到很多钱。

Japan’s own media perceived overinvestment in the semiconductor sector, with newspaper headlines warning of “reckless investment competition” and “investment they cannot stop.” CEOs of Japan’s memory chip producers couldn’t bring themselves to stop building new chip fabs, even if they weren’t profitable. “If you start worrying” about overinvestment, one Hitachi executive admitted, “you can’t sleep at night.” So long as banks kept lending, it was easier for CEOs to keep spending than to admit they had no path to profitability. America’s arm’s-length capital markets hadn’t felt like an advantage in the 1980s, but the risk of losing financing helped keep American firms on their toes. Japanese DRAM makers would have benefitted from Andy Grove’s paranoia or Jack Simplot’s wisdom about commodity market volatility. Instead, they all poured investment into the same market, guaranteeing that few made much money.

索尼在日本半导体公司中独树一帜,从未大力押注 DRAM,它成功开发了创新的新产品,例如用于图像传感器的专用芯片。当光子撞击它们的硅时,这些芯片会产生与光强相关的电荷,让芯片将图像转换为数字数据。因此,索尼完全有能力引领数码相机革命,该公司的图像感知芯片今天仍然是世界一流的。即便如此,该公司未能削减对亏损部门的投资,其盈利能力从 1990 年代初开始下滑。

Sony, which was unique among Japanese semiconductor firms in never betting heavily on DRAMs, succeeded in developing innovative new products, like specialized chips for image sensors. When photons strike their silicon, these chips create electric charges that are correlated to the strength of the light, letting the chips convert images into digital data. Sony was therefore well placed to lead the digital camera revolution, and the company’s chips that sense images today remain world-class. Even still, the company failed to cut investment in loss-making segments, and its profitability slumped beginning in the early 1990s.

然而,大多数日本大型 DRAM 生产商未能利用其在 1980 年代的影响力来推动创新。在 DRAM 巨头东芝,一位名叫 Fujio Masuoka 的中层工厂经理在 1981 年开发了一种新型存储芯片,它与 DRAM 不同,即使在断电后也可以继续“记住”数据。东芝忽略了这一发现,所以是英特尔带来了这种新型存储芯片,通常称为“闪存”或 NAND,即将上市。

Most of Japan’s big DRAM producers, however, failed to take advantage of their influence in the 1980s to drive innovation. At Toshiba, a DRAM giant, a mid-ranking factory manager named Fujio Masuoka developed a new type of memory chip in 1981 that, unlike DRAM, could continue “remembering” data even after it was powered off. Toshiba ignored this discovery, so it was Intel that brought this new type of memory chip, commonly called “flash” or NAND, to market.

然而,日本芯片公司犯下的最大错误是错过了个人电脑的崛起。没有一家日本芯片巨头能够复制英特尔转向微处理器或对 PC 生态系统的掌控。只有一家日本公司,NEC,真正尝试过,但它从未赢得过微控制器市场的一小部分份额。对于安迪·格鲁夫和英特尔来说,在微处理器上赚钱是生死攸关的大事。日本的 DRAM 公司拥有巨大的市场份额和很少的资金限制,忽略了微处理器市场,直到为时已晚。结果,个人电脑革命主要使美国芯片公司受益。到日本股市崩盘时,日本的半导体主导地位已经被削弱。1993年,美国重新夺回半导体出货量第一名。1998年,韩国企业超越日本成为全球最大的DRAM生产国,到 1998 年为 20%。

The biggest error that Japan’s chip firms made, however, was to miss the rise of PCs. None of the Japanese chip giants could replicate Intel’s pivot to microprocessors or its mastery of the PC ecosystem. Only one Japanese firm, NEC, really tried, but it never won more than a tiny share of the microprocessor market. For Andy Grove and Intel, making money on microprocessors was a matter of life or death. Japan’s DRAM firms, with massive market share and few financial constraints, ignored the microprocessor market until it was too late. As a result, the PC revolution mostly benefitted American chip firms. By the time Japan’s stock market crashed, Japan’s semiconductor dominance was already eroding. In 1993, the U.S. retook first place in semiconductor shipments. In 1998, South Korean firms had overtaken Japan as the world’s largest producers of DRAM, while Japan’s market share fell from 90 percent in the late 1980s to 20 percent by 1998.

日本的半导体野心为该国不断扩大的全球地位奠定了基础,但这个基础现在看起来很脆弱。在《可以说不的日本》中,石原和森田认为日本可以利用芯片优势对美国和苏联施加权力。但是当战争终于来了,在波斯湾这个意想不到的舞台上,美国军队可能让大多数观察者感到震惊。在数字时代的第一场战争中,日本拒绝加入向海湾地区派遣军队以将伊拉克军队驱逐出科威特的 28 个国家。相反,东京通过发送支票来支付联军的费用并参与其中支持伊拉克的邻国。当美国宝石路激光制导炸弹袭击伊拉克坦克纵队时,这种金融外交看起来无能为力。

Japan’s semiconductor ambitions had underwritten the country’s expanding sense of its global position, but this foundation now looked brittle. In The Japan That Can Say No, Ishihara and Morita had argued Japan could use chip dominance to exert power over both the United States and the USSR. But when war finally came, in the unexpected arena of the Persian Gulf, American military might astounded most observers. In the first war of the digital era, Japan declined to join the twenty-eight countries that sent troops to the Gulf to eject Iraqi forces from Kuwait. Instead, Tokyo participated by sending checks to pay for coalition armies and to support Iraq’s neighbors. As American Paveway laser-guided bombs pummeled Iraqi tank columns, this financial diplomacy looked impotent.

森田在 1993 年中风,导致身体虚弱。他从公众视野中撤退,并在夏威夷度过了余生,直到 1999 年去世。森田的合著者石原一直坚持认为日本需要在世界舞台上站稳脚跟。就像破纪录一样,他在 1994 年出版了《可以说不的亚洲》,随后又出版了《可以再次说不的日本》几年后。但对大多数日本人来说,石原的论点不再有意义。在 1980 年代,他预测芯片将塑造军事平衡并定义技术的未来是正确的。但他认为这些芯片将在日本制造是错误的。面对美国的复兴,该国的半导体公司在 1990 年代一直在萎缩。日本挑战美国霸权的技术基础开始崩溃。

Morita suffered a stroke in 1993 that caused debilitating health problems. He retreated from public view and spent most of the remainder of his life in Hawaii, before dying in 1999. Morita’s coauthor, Ishihara, kept insisting that Japan needed to assert itself on the world stage. Like a broken record, he published The Asia That Can Say No in 1994 followed by The Japan That Can Say No Again several years later. But to most Japanese, Ishihara’s argument no longer made sense. In the 1980s, he’d been right to predict chips would shape the military balance and define the future of technology. But he was wrong to think those chips would be made in Japan. The country’s semiconductor firms spent the 1990s shrinking in the face of America’s resurgence. The technological basis for Japan’s challenge to American hegemony began to crumble.

与此同时,美国唯一的另一个严重挑战者正在走向崩溃。1990 年,苏联领导人米哈伊尔·戈尔巴乔夫(Mikhail Gorbachev)在认识到通过指挥方法和“复制”战略克服技术落后的努力已无望后,抵达硅谷进行正式访问。这座城市的科技大亨们为他准备了一场沙皇般的盛宴。大卫帕卡德和苹果公司的史蒂夫沃兹尼亚克坐在戈尔巴乔夫旁边,享用美酒佳肴。戈尔巴乔夫毫不掩饰他选择访问加州湾区的原因。“明天的想法和技术诞生于加利福尼亚,”他在斯坦福大学的一次演讲中宣称。这正是奥加科夫元帅十多年来一直警告他的苏联领导人的事情。

The only other serious challenger to the United States, meanwhile, was careening toward collapse. In 1990, having recognized that efforts to overcome technological backwardness via command methods and the “copy it” strategy were hopeless, Soviet leader Mikhail Gorbachev arrived in Silicon Valley for an official visit. The city’s tech tycoons treated him with a feast fit for a tsar. David Packard and Apple’s Steve Wozniak sat alongside Gorbachev as he was wined and dined. Gorbachev made no secret of why he chose to visit California’s Bay Area. “The ideas and technologies of tomorrow are born here in California,” he declared in a speech at Stanford. This was exactly what Marshal Ogarkov had been warning his fellow Soviet leaders of for over a decade.

戈尔巴乔夫承诺通过从东欧撤军来结束冷战,他希望获得美国技术作为交换。他与美国的科技高管会面,鼓励他们投资苏联。当戈尔巴乔夫访问斯坦福大学时,他在校园里走来走去时向观众击掌。“冷战现在已经过去了,”这位苏联领导人在斯坦福大学对听众说。“我们不要争论谁赢了。”

Gorbachev promised to end the Cold War by withdrawing Soviet troops from Eastern Europe, and he wanted access to American technologies in exchange. Meeting with America’s tech executives, he encouraged them to invest in the USSR. When Gorbachev visited Stanford University, he high-fived spectators as he walked around campus. “The Cold War is now behind us,” the Soviet leader told an audience at Stanford. “Let’s not wrangle over who won it.”

但很明显谁赢了,为什么赢了。奥加科夫十年前就发现了这种动态,尽管当时他希望苏联可能战胜它。和其他苏联军事领导层一样,随着时间的推移,他变得更加悲观。早在 1983 年,奥加科夫甚至私下告诉美国记者 Les Gelb,“冷战结束了,你赢了。” 苏联的火箭一如既往的强大。它拥有世界上最大的核武库。但是它的半导体生产跟不上,它的计算机产业落后,它的通信和监视技术落后,军事后果是灾难性的。“所有现代军事能力都基于经济创新、技术和经济实力,”奥加科夫向盖尔布解释道。“军事技术是基于计算机的。您在计算机方面远远领先于我们…… 在你的国家,每一个小孩5岁就有电脑了。”

But it was obvious who won, and why. Ogarkov had identified the dynamic a decade earlier, though at the time he hoped the USSR might overcome it. Like the rest of the Soviet military leadership, he’d grown more pessimistic over time. As early as 1983, Ogarkov had gone so far as to tell American journalist Les Gelb—off the record—that “the Cold War is over and you have won.” The Soviet Union’s rockets were as powerful as ever. It had the world’s largest nuclear arsenal. But its semiconductor production couldn’t keep up, its computer industry fell behind, its communications and surveillance technologies lagged, and the military consequences were disastrous. “All modern military capability is based on economic innovation, technology, and economic strength,” Ogarkov explained to Gelb. “Military technology is based on computers. You are far, far ahead of us with computers…. In your country, every little child has a computer from age 5.”

在轻松击败萨达姆·侯赛因的伊拉克之后,每个人都可以看到美国巨大的新战斗力。这给苏联军队和克格勃造成了危机,他们既尴尬又不敢承认他们是多么果断地被打败了。安全负责人领导了一场针对戈尔巴乔夫的士气低落的政变企图,三天后爆发。对于一个曾经强大的国家来说,这是一个可悲的结局,它无法接受其军事实力的痛苦衰落。俄罗斯芯片行业也面临着自己的耻辱,1990 年代,一家工厂缩减为为麦当劳生产微型芯片开心乐园餐玩具。冷战结束了;硅谷赢了。

After the easy defeat of Saddam Hussein’s Iraq, America’s vast new fighting power was visible to everyone. This caused a crisis in the Soviet military and the KGB, who were embarrassed yet afraid to admit how decisively they were outgunned. The security chiefs led a demoralized coup attempt against Gorbachev that sputtered out after three days. It was a pathetic end for a once-powerful country, which couldn’t come to terms with the painful decline in its military power. The Russian chip industry faced humiliation of its own, with one fab reduced in the 1990s to producing tiny chips for McDonald’s Happy Meal toys. The Cold War was over; Silicon Valley had won.

第五部分 集成电路,集成世界?

PART V INTEGRATED CIRCUITS, INTEGRATED WORLD?

第 29 章 “我们希望台湾有半导体产业”

CHAPTER 29 “We Want a Semiconductor Industry in Taiwan”

1985年,台湾有权势的部长KT Li将张忠谋召入他在台北的办公室。自从李帮助说服德州仪器公司在岛上建立第一家半导体工厂以来,已经过去了将近 20 年。从那以后的二十年里,李与德州仪器的领导人建立了密切的联系,每次在美国时都会拜访 Pat Haggerty 和 Morris Chang,并说服其他电子公司效仿德州仪器并在台湾开设工厂。1985年,他聘请张先生领导台湾芯片业。“我们想在台湾促进半导体产业,”他告诉张。“告诉我,”他继续说,“你需要多少钱。”

In 1985, Taiwan’s powerful minister K. T. Li called Morris Chang into his office in Taipei. Nearly two decades had passed since Li had helped convince Texas Instruments to build its first semiconductor facility on the island. In the twenty years since then, Li had forged close ties with Texas Instrument’s leaders, visiting Pat Haggerty and Morris Chang whenever he was in the U.S. and convincing other electronics firms to follow TI and open factories in Taiwan. In 1985, he hired Chang to lead Taiwan’s chip industry. “We want to promote a semiconductor industry in Taiwan,” he told Chang. “Tell me,” he continued, “how much money you need.”

1990 年代是“全球化”这个词第一次被普遍使用的年代,尽管自飞兆半导体成立之初,芯片行业就依赖于国际生产和组装。自 1960 年代以来,台湾有意将自己纳入半导体供应链,作为提供就业、获取先进技术和加强与美国安全关系的战略。1990 年代,在台积电 164Manufacturing 惊人崛起的推动下,台湾的重要性开始增长在台湾政府的大力支持下,Chang创立的公司。

The 1990s were the years when the word “globalization” first became commonly used, though the chip industry had relied on international production and assembly since the earliest days of Fairchild Semiconductor. Taiwan had deliberately inserted itself into semiconductor supply chains since the 1960s, as a strategy to provide jobs, acquire advanced technology, and to strengthen its security relationship with the United States. In the 1990s, Taiwan’s importance began to grow, driven by the spectacular rise of the Taiwan Semiconductor 164Manufacturing Company, which Chang founded with strong backing from the Taiwanese government.

当张在 1985 年被台湾政府聘用领导该国卓越的电子研究所时,台湾是亚洲组装半导体设备的领导者之一——采用国外制造的芯片,对其进行测试,并将它们连接到塑料或陶瓷封装上。台湾政府曾尝试通过从美国 RCA 授权半导体制造技术并于 1980 年成立一家名为 UMC 的芯片制造商来打入芯片制造行业,但该公司的能力远远落后于尖端技术。台湾拥有大量半导体行业的工作岗位,但只获得了一小部分利润,因为芯片行业的大部分资金来自设计和生产最先进芯片的公司。像李部长这样的官员知道,只有超越简单地组装在其他地方设计和制造的部件,该国的经济才能继续增长。

When Chang was hired by Taiwan’s government in 1985 to lead the country’s preeminent electronics research institute, Taiwan was one of Asia’s leaders in assembling semiconductor devices—taking chips made abroad, testing them, and attaching them to plastic or ceramic packages. Taiwan’s government had tried breaking into the chipmaking business by licensing semiconductor manufacturing technology from America’s RCA and founding a chipmaker called UMC in 1980, but the company’s capabilities lagged far behind the cutting edge. Taiwan boasted plenty of semiconductor industry jobs, but captured only a small share of the profit, since most money in the chip industry was made by firms designing and producing the most advanced chips. Officials like Minister Li knew the country’s economy would keep growing only if it advanced beyond simply assembling components designed and fabricated elsewhere.

1968 年张忠谋首次访问台湾时,该岛正与香港、韩国、新加坡和马来西亚竞争。现在三星和韩国的其他大型企业集团正在向最先进的存储芯片投入资金。新加坡和马来西亚正试图复制韩国从组装半导体到制造半导体的转变,尽管成功率不及三星。台湾必须不断提高能力,以维持其在半导体供应链底层的地位。

When Morris Chang had first visited Taiwan in 1968, the island was competing with Hong Kong, South Korea, Singapore, and Malaysia. Now Samsung and South Korea’s other big conglomerates were pouring funds into the most advanced memory chips. Singapore and Malaysia were trying to replicate South Korea’s shift from assembling semiconductors to fabricating them, though with less success than Samsung. Taiwan had to improve its capabilities constantly simply to maintain its position in the bottom rungs of the semiconductor supply chain.

最大的威胁是中华人民共和国。在台湾海峡对岸,毛泽东于 1976 年去世,减少了迫在眉睫的入侵威胁。但中国现在构成了经济挑战。在其新的、后毛泽东时代的领导下,中国开始通过吸引台湾用来摆脱贫困的一些基本制造和装配工作来融入全球经济。由于工资较低和数亿农民渴望以自给农业换取工厂工作,中国进入电子组装行业有可能让台湾破产。相当于经济“战争”,台湾官员向来访的德州仪器高管抱怨。在价格上与中国竞争是不可能的。台湾必须自己生产先进技术。

The biggest threat was the People’s Republic of China. Across the Taiwan Strait, Mao Zedong had died in 1976, reducing the threat of imminent invasion. But China now posed an economic challenge. Under its new, post-Mao leadership, China began integrating into the global economy by attracting some of the basic manufacturing and assembly jobs that Taiwan had used to lift itself out of poverty. With lower wages and several hundred million peasants eager to trade subsistence farming for factory jobs, China’s entry into electronics assembly threatened to put Taiwan out of business. It amounted to economic “warfare,” Taiwanese officials complained to visiting Texas Instruments executives. It was impossible to compete with China on price. Taiwan had to produce advanced technology itself.

KT Li 求助于首先帮助将半导体组装带到台湾的人:Morris Chang。在德州仪器工作了 20 多年后,Chang 于 1980 年代初离开了公司,因为他的 CEO 职位被忽略了,“放牧,”他后来说。他花了一年时间在纽约经营一家名为 General Instrument 的电子公司,但不久后辞职,对工作不满意。他亲自帮助建立了世界半导体产业。TI 的超高效制造工艺是他在提高产量方面的实验和专业知识的结果。他在 TI 想要的工作——CEO——将使他成为芯片行业的佼佼者,与 Bob Noyce 或 Gordon Moore 相提并论。因此,当台湾政府打来电话,提议让他负责台湾的芯片产业,并提供一张空白支票来资助他的计划时,Chang 觉得这个提议很有趣。五十四岁时,他正在寻找新的挑战。

K. T. Li turned to the person who’d first helped bring semiconductor assembly to Taiwan: Morris Chang. After over two decades with Texas Instruments, Chang had left the company in the early 1980s after being passed over for the CEO job and “put out to pasture,” he’d later say. He spent a year running an electronics company in New York called General Instrument, but resigned soon after, dissatisfied with the work. He’d personally helped build the world’s semiconductor industry. TI’s ultra-efficient manufacturing processes were the result of his experimentation and expertise in improving yields. The job he’d wanted at TI—CEO—would have placed him at the top of the chip industry, on par with Bob Noyce or Gordon Moore. So when the government of Taiwan called, offering to put him in charge of the island’s chip industry and providing a blank check to fund his plans, Chang found the offer intriguing. At age fifty-four, he was looking for a new challenge.

尽管大多数人都说张“回归”台湾,但他与台湾最紧密的联系是他帮助建立的德州仪器设施,以及台湾声称是张长大的中国大陆的合法政府,但自从近四年前逃离后,他就再也没有去过那里。到 1980 年代中期,Chang 居住时间最长的地方是德克萨斯州。他在 TI 获得了国防相关工作的美国安全许可。他可以说是德州人多于台湾人。“台湾是对我来说是个陌生的地方,”他后来回忆道。

Though most people speak of Chang “returning” to Taiwan, his strongest connection to the island was the Texas Instruments facilities that he helped establish, and by Taiwan’s claim to be the legitimate government of China, the country that Chang grew up in, but that he hadn’t visited since fleeing nearly four decades earlier. By the mid-1980s, the place Chang had lived the longest was Texas. He held a U.S. security clearance for defense-related work at TI. He was arguably more Texan than Taiwanese. “Taiwan was a strange place to me,” he’d later recall.

然而,建设台湾的半导体产业听起来像是一个令人兴奋的挑战。领导台湾政府工业技术研究所,张正式获得的职位将使他成为台湾芯片开发工作的中心。政府融资的承诺使交易更加甜蜜。实际上负责岛上的半导体部门保证 Chang 不需要回答任何人,除了像 KT Li 这样的部长,他们承诺给他宽阔的余地。德州仪器从来没有像这样开过空白支票。张知道他需要很多钱,因为他的商业计划是基于一个激进的想法。如果它奏效,它将颠覆电子行业,让他和台湾控制世界上最先进的技术。

However, building Taiwan’s semiconductor industry sounded like an exciting challenge. Directing the Taiwanese government’s Industrial Technology Research Institute, the position that Chang was formally offered, would place him at the center of Taiwan’s chip development efforts. The promise of government financing sweetened the deal. Being placed de facto in charge of the island’s semiconductor sector guaranteed Chang wouldn’t have to answer to anyone except ministers like K. T. Li, who promised to give him wide leeway. Texas Instruments never handed out blank checks like this. Chang knew he’d need a lot of money, because his business plan was based on a radical idea. If it worked, it would upend the electronics industry, placing him—and Taiwan—in control of the world’s most advanced technology.

早在 1970 年代中期,还在 TI 工作时,Chang 就曾想过创建一家半导体公司,生产由客户设计的芯片。当时,像 TI、英特尔和摩托罗拉这样的芯片公司大多生产他们自己设计的芯片。1976 年 3 月,Chang 向 TI 的高管们介绍了这种新的商业模式。“计算能力的低成本,”他向 TI 同事解释说,“将会打开大量现在没有半导体服务的应用,”为芯片创造了新的需求来源,芯片很快将用于从手机到汽车再到洗碗机的所有领域。他推断,制造这些产品的公司缺乏生产半导体的专业知识,因此他们更愿意将制造外包给专家。此外,随着技术的进步和晶体管的缩小,制造设备和研发的成本将会上升。只有生产大量芯片的公司才有成本竞争力。

As early as the mid-1970s, while still at TI, Chang had toyed with the idea of creating a semiconductor company that would manufacture chips designed by customers. At the time, chip firms like TI, Intel, and Motorola mostly manufactured chips they had designed in-house. Chang pitched this new business model to fellow TI executives in March 1976. “The low cost of computing power,” he explained to his TI colleagues, “will open up a wealth of applications that are not now served by semiconductors,” creating new sources of demand for chips, which would soon be used in everything from phones to cars to dishwashers. The firms that made these goods lacked the expertise to produce semiconductors, so they’d prefer to outsource fabrication to a specialist, he reasoned. Moreover, as technology advanced and transistors shrank, the cost of manufacturing equipment and R&D would rise. Only companies that produced large volumes of chips would be cost-competitive.

TI 的其他高管并不相信。当时,1976 年,还没有任何“无晶圆厂”公司设计芯片但缺乏自己的晶圆厂,尽管 Chang 预测这样的公司很快就会出现。德州仪器已经赚了很多钱,所以在尚不存在的市场上赌博似乎有风险。这个想法被悄悄地搁置了。

TI’s other executives weren’t convinced. At the time, in 1976, there weren’t any “fabless” companies that designed chips but lacked their own fabs, though Chang predicted such companies would soon emerge. Texas Instruments was already making plenty of money, so gambling on markets that didn’t yet exist seemed risky. The idea was quietly binned.

Chang从未忘记铸造理念。他认为随着时间的推移它正在成熟,特别是在 Lynn Conway 和 Carver Mead 在芯片设计方面的革命使芯片设计与制造分离变得容易得多之后,他们认为这将为半导体创造古腾堡时刻。

Chang never forgot the foundry concept. He thought it was ripening as time passed, particularly after Lynn Conway and Carver Mead’s revolution in chip design made it far easier to separate chip design from manufacturing, which they thought would create a Gutenberg moment for semiconductors.

在台湾,岛上的一些电气工程师也有类似的想法。曾协助经营台湾工业技术研究院的施正泰曾邀请米德访台1980 年代中期,他分享了他对古腾堡半导体的愿景。因此,将芯片设计和制造分离的想法在李嘉诚部长向张忠谋提供空白支票以建立台湾芯片产业之前,已经在台湾渗透了几年。

In Taiwan, some of the island’s electrical engineers were thinking along similar lines. Chintay Shih, who helped run Taiwan’s Industrial Technology Research Institute, had invited Mead to visit Taiwan in the mid-1980s to share his vision of Gutenberg for semiconductors. The idea of separating chip design and manufacturing had therefore already been percolating in Taiwan for several years before Minister K. T. Li offered Morris Chang a blank check to build Taiwan’s chip industry.

李部长兑现了他的承诺,为常制定的商业计划筹集资金。台湾政府为台积电提供了 48% 的启动资金,只是规定常务要找一家外国芯片公司提供先进的生产技术。他被他在 TI 的前同事和英特尔拒绝了。“莫里斯,你这个时代有很多好主意,”戈登摩尔告诉他。“这不是其中之一。” 然而,Chang 说服荷兰半导体公司飞利浦出资 5800 万美元,转让其生产技术,并许可知识产权以换取台积电27.5%的股份。

Minister Li followed through on his promise to find the money for the business plan Chang drew up. The Taiwanese government provided 48 percent of the startup capital for TSMC, stipulating only that Chang find a foreign chip firm to provide advanced production technology. He was turned down by his former colleagues at TI and by Intel. “Morris, you’ve had a lot of good ideas in your time,” Gordon Moore told him. “This isn’t one of them.” However, Chang convinced Philips, the Dutch semiconductor company, to put up $58 million, transfer its production technology, and license intellectual property in exchange for a 27.5 percent stake in TSMC.

其余资金来自台湾富人,他们被政府“要求”投资。“通常发生的情况是,政府的一位部长会打电话给台湾的商人,”张解释说,“让他投资。” 政府要求岛上几个最富有的家庭提供资金,这些家庭拥有专门从事塑料、纺织品和化学品的公司。当一位商人在与张三会面后拒绝投资时,台湾总理打电话给这位吝啬的行政人员并提醒他:“过去二十年来,政府对你一直很好。你现在最好为政府做点什么。” Chang的芯片代工厂的支票很快就到了。政府还为台积电提供了丰厚的税收优惠,确保该公司有足够的资金进行投资。从第一天开始,台湾政府的一个项目。

The rest of the capital was raised from wealthy Taiwanese who were “asked” by the government to invest. “What generally happened was that one of the ministers in the government would call a businessman in Taiwan,” Chang explained, “to get him to invest.” The government asked several of the island’s wealthiest families, who owned firms that specialized in plastics, textiles, and chemicals, to put up the money. When one businessman declined to invest after three meetings with Chang, Taiwan’s prime minister called the stingy executive and reminded him, “The government has been very good to you for the last twenty years. You better do something for the government now.” A check for Chang’s chip foundry arrived soon after. The government also provided generous tax benefits for TSMC, ensuring the company had plenty of money to invest. From day one, TSMC wasn’t really a private business: it was a project of the Taiwanese state.

台积电早期成功的一个关键因素是与美国芯片产业有着深厚的联系。它的大部分客户都是美国芯片设计师,许多顶级员工曾在硅谷工作过。Morris Chang 聘请了另一位德州仪器 (TI) 前高管 Don Brooks 担任1991 年至 1997 年担任台积电总裁。“向我汇报的大多数人,下两级,”布鲁克斯回忆道,“都曾在美国工作过……他们都曾为摩托罗拉、英特尔或 TI 工作过。” 在 1990 年代的大部分时间里,台积电一半的销售额是卖给美国公司的。与此同时,该公司的大多数高管都在美国大学接受过顶尖博士课程的培训。

A crucial ingredient in TSMC’s early success was deep ties with the U.S. chip industry. Most of its customers were U.S. chip designers, and many top employees had worked in Silicon Valley. Morris Chang hired Don Brooks, another former Texas Instruments executive, to work as TSMC’s president from 1991 to 1997. “Most of the guys who reported to me, down two levels,” Brooks recalled, “all had some experience in the U.S… they all worked for Motorola, Intel, or TI.” Throughout much of the 1990s, half of TSMC’s sales were to American companies. Most of the company’s executives, meanwhile, trained in top doctoral programs at U.S. universities.

这种共生惠及台湾和硅谷。在台积电之前,几家主要位于硅谷的小公司曾尝试围绕芯片设计开展业务,通过外包制造来避免建立自己的晶圆厂的成本。这些“无晶圆厂”公司有时能够说服拥有闲置产能的更大芯片制造商来制造他们的芯片。然而,在更大的芯片制造商自己的生产计划之后,它们始终处于二流地位。更糟糕的是,他们面临着制造合作伙伴窃取他们创意的持续风险。此外,他们必须驾驭每个大型芯片制造商略有不同的制造流程。不必建造晶圆厂大大降低了启动成本,但依靠竞争对手制造芯片始终是一种冒险的商业模式。

This symbiosis benefitted Taiwan and Silicon Valley. Before TSMC, a couple of small companies, mostly based in Silicon Valley, had tried building businesses around chip design, avoiding the cost of building their own fabs by outsourcing the manufacturing. These “fabless” firms were sometimes able to convince a bigger chipmaker with spare capacity to manufacture their chips. However, they always had second-class status behind the bigger chipmakers’ own production plans. Worse, they faced the constant risk that their manufacturing partners would steal their ideas. In addition, they had to navigate manufacturing processes that were slightly different at each big chipmaker. Not having to build fabs dramatically reduced startup costs, but counting on competitors to manufacture chips was always a risky business model.

台积电的成立给了所有芯片设计者一个可靠的合作伙伴。Chang 承诺永远不会设计芯片,只会制造它们。台积电不与客户竞争;如果他们这样做了,它就成功了。十年前,Carver Mead 预言了芯片制造的古腾堡时刻,但有一个关键的区别。旧的德国印刷商试图建立对印刷的垄断,但未能成功。他无法阻止他的技术在欧洲迅速传播,使作家和印刷厂都受益。

The founding of TSMC gave all chip designers a reliable partner. Chang promised never to design chips, only to build them. TSMC didn’t compete with its customers; it succeeded if they did. A decade earlier, Carver Mead had prophesied a Gutenberg moment in chipmaking, but there was one key difference. The old German printer had tried and failed to establish a monopoly over printing. He couldn’t stop his technology from quickly spreading across Europe, benefitting authors and print shops alike.

在芯片行业,通过降低启动成本,Chang 的代工模式催生了数十家新的“作者”——无晶圆厂芯片设计公司——通过将计算能力应用于各种设备,从而改变了科技行业。然而,作者身份的民主化恰逢数字印刷机的垄断。芯片制造的经济性需要不断整合。生产最多芯片的公司具有内在优势,169 提高了产量并将资本投资成本分摊给更多客户。台积电的业务在 1990 年代蓬勃发展,其制造工艺不断改进。莫里斯·张想成为数字时代的古腾堡。他最终变得更加强大。当时几乎没有人意识到这一点,但张,台积电,

In the chip industry, by lowering startup costs, Chang’s foundry model gave birth to dozens of new “authors”—fabless chip design firms—that transformed the tech sector by putting computing power in all sorts of devices. However, the democratization of authorship coincided with a monopolization of the digital printing press. The economics of chip manufacturing required relentless consolidation. Whichever company produced the most chips had a built-in advantage, 169improving its yield and spreading capital investment costs over more customers. TSMC’s business boomed during the 1990s and its manufacturing processes improved relentlessly. Morris Chang wanted to become the Gutenberg of the digital era. He ended up vastly more powerful. Hardly anyone realized it at the time, but Chang, TSMC, and Taiwan were on a path toward dominating the production of the world’s most advanced chips.

第 30 章 “所有人都必须制造半导体”

CHAPTER 30 “All People Must Make Semiconductors”

1987 年,也就是张忠谋创立台积电的同一年,在西南数百英里处,一位当时名不见经传的工程师任正非成立了一家名为华为的电子贸易公司。台湾是个小岛,野心很大。它不仅与世界上最先进的芯片公司有着深厚的联系,而且与数千名在斯坦福和伯克利等大学接受过教育的工程师有着深厚的联系。相比之下,中国人口众多,但贫穷且技术落后。然而,一项新的经济开放政策导致贸易繁荣,特别是通过香港进口或走私货物。华为的创立地深圳就坐落在边境对面。

In 1987, the same year that Morris Chang founded TSMC, several hundred miles to the southwest a then-unknown engineer named Ren Zhengfei established an electronics trading company called Huawei. Taiwan was a small island with big ambitions. It had deep connections not just with the world’s most advanced chip companies but also thousands of engineers who’d been educated at universities like Stanford and Berkeley. China, by contrast, had a vast population but was impoverished and technologically backward. A new policy of economic openness had caused trade to boom, however, particularly via Hong Kong, through which goods could be imported or smuggled. Shenzhen, where Huawei was founded, sat just across the border.

在台湾,张忠谋将目光投向了打造一些世界上最先进的芯片,并赢得硅谷巨头的青睐。在深圳,任正非在香港买了便宜的电信设备,然后在全国以更高的价格出售。他交易的设备使用集成电路,但自己生产芯片的想法似乎很荒谬。1980 年代,以电子工业部部长、后来的中国国家主席江泽民为首的中国政府将电子产品确定为优先。当时,中国国内生产的最先进、最广泛使用的芯片是 DRAM与英特尔在 1970 年代初投放市场的第一款 DRAM 相同的存储容量,使中国落后于前沿十多年。

In Taiwan, Morris Chang set his sights on building some of the world’s most advanced chips and winning Silicon Valley giants as his customers. In Shenzhen, Ren Zhengfei bought cheap telecommunications equipment in Hong Kong and sold it for a higher price across China. The equipment he traded used integrated circuits, but the idea of producing his own chips would have seemed absurd. In the 1980s, the Chinese government, led by minister of the electronics industry and later president of China Jiang Zemin, identified electronics as a priority. At the time, the most advanced, widely used chip that China produced domestically was a DRAM with roughly the same storage capacity as the first DRAM Intel had brought to market in the early 1970s, putting China over a decade behind the cutting edge.

如果不是共产党统治,中国可能在半导体行业中发挥了更大的作用。发明集成电路时,中国拥有许多帮助日本、台湾和韩国吸引美国半导体投资的要素,例如庞大的低成本劳动力和受过良好教育的科学精英。然而,在 1949 年夺取政权后,共产党人怀疑地看待与外国的联系。对于张忠谋这样的人来说,在斯坦福完成学业后回到中国,就意味着一定的贫困,可能会被监禁或死亡。许多在革命前从中国大学毕业的最优秀的毕业生最终在台湾或加利福尼亚工作,为中国的主要竞争对手打造电子产品能力。

Were it not for Communist rule, China might have played a much larger role in the semiconductor industry. When the integrated circuit was invented, China had many of the ingredients that helped Japan, Taiwan, and South Korea attract American semiconductor investment, like a vast, low-cost workforce and a well-educated scientific elite. However, after seizing power in 1949, the Communists looked at foreign connections with suspicion. For someone like Morris Chang, returning to China after finishing his studies at Stanford would have meant certain poverty and possible imprisonment or death. Many of the best graduates from China’s universities before the revolution ended up working in Taiwan or in California, building the electronics capabilities of the PRC’s primary rivals.

与此同时,中国共产党政府犯了与苏联相同的错误,尽管形式更为极端。早在 1950 年代中期,北京已将半导体设备确定为科学优先事项。很快,他们呼吁北京大学和其他科学中心的研究人员发挥技能——包括一些在革命前在伯克利、麻省理工学院、哈佛或普渡大学接受过培训的科学家。到 1960 年,中国在北京成立了第一个半导体研究所。大约在同一时间,该国开始制造简单的晶体管收音机。1965 年,在 Bob Noyce 和 Jack Kilby 五年之后,中国工程师打造了他们的第一个集成电路。

China’s Communist government, meanwhile, made the same mistakes the Soviet Union did, though in more extreme forms. As early as the mid-1950s, Beijing had identified semiconductor devices as a scientific priority. Soon, they were calling on the skills of researchers at Peking University and other scientific centers—including some scientists who’d been trained before the revolution at Berkeley, MIT, Harvard, or Purdue. By 1960, China had established its first semiconductor research institute, in Beijing. Around the same time, the country began manufacturing simple transistor radios. In 1965, Chinese engineers forged their first integrated circuit, a half decade after Bob Noyce and Jack Kilby.

然而,毛泽东的激进主义使其无法吸引外国投资或开展严肃的科学工作。在中国生产出第一条集成电路后的第二年,毛泽东让中国陷入文化大革命,认为专业知识是破坏社会主义平等的特权来源。毛泽东的游击队对国家的教育体系发动了战争。数以千计的科学家和专家被派往贫困村庄当农民。许多其他人被简单地杀死了。毛主席“1968年7月21日发出的光辉指示”坚持“必须缩短学制,改革教育,推行无产阶级政治…… 从有实践经验的工农中选拔学生,重返生产经过几年的学习。”

However, Mao’s radicalism made it impossible to attract foreign investment or conduct serious science. The year after China produced its first integrated circuit, Mao plunged the country into the Cultural Revolution, arguing that expertise was a source of privilege that undermined socialist equality. Mao’s partisans waged war on the country’s educational system. Thousands of scientists and experts were sent to work as farmers in destitute villages. Many others were simply killed. Chairman Mao’s “Brilliant Directive issued on July 21, 1968” insisted that “it is essential to shorten the length of schooling, revolutionize education, put proletarian politics in command…. Students should be selected from among workers and peasants with practical experience, and they should return to production after a few years study.”

用受过良好教育的员工建立先进产业的想法是荒谬的。毛泽东努力将外国技术和思想拒之门外,更是如此。美国的限制阻止了中国购买先进的半导体设备,但毛泽东增加了他自己的禁运。他想要完全自力更生,并指责他的政治对手试图用外国零件感染中国的芯片产业,尽管中国自己不能生产很多先进的元件。他的宣传机器敦促支持“惊天动地的群众运动……电子工业的独立和自力更生的发展。”

The idea of building advanced industries with poorly educated employees was absurd. Even more so was Mao’s effort to keep out foreign technology and ideas. U.S. restrictions prevented China from buying advanced semiconductor equipment, but Mao added his own self-imposed embargo. He wanted complete self-reliance and accused his political rivals of trying to infect China’s chip industry with foreign parts, even though China couldn’t produce many advanced components itself. His propaganda machine urged support for “the earth-shaking mass movement for the… independent and self-reliant development of the electronic industry.”

毛不仅仅是对外国芯片持怀疑态度;有时他担心所有电子产品本质上都是反社会主义的。他的政治对手刘少奇赞同“现代电子技术”将“为我们的工业带来巨大飞跃”和“使中国成为第一个拥有一流电子技术的社会主义新兴工业化强国”的想法。总是将社会主义与烟囱联系在一起的毛泽东抨击了这个想法。毛的一位支持者认为,将电子产品视为未来是“反动的”,而在建立一个“只有钢铁工业应该起主导作用”的情况下,这是显而易见的。中国的社会主义乌托邦。

Mao wasn’t simply skeptical of foreign chips; at times he worried that all electronic goods were intrinsically anti-socialist. His political rival Liu Shaoqi had endorsed the idea that “modern electronic technology” would “bring about a big leap forward for our industry” and would “make China the first newly industrialized socialist power with first-rate electronic technology.” Mao, who always associated socialism with smokestacks, attacked the idea. It was “reactionary,” one of Mao’s supporters argued, to see electronics as the future, when it was obvious that “only the iron and steel industry should play a leading role” in building a socialist utopia in China.

在 1960 年代,毛泽东赢得了围绕中国半导体产业的政治斗争,淡化其重要性并切断了与外国技术的联系。大多数中国科学家对主席将他们送到农民农场去研究无产阶级政治而不是半导体工程,从而毁了他们的研究和他们的生活感到不满。一位被派往农村的中国著名光学专家在农村再教育中幸存下来,他以粗粮、煮白菜和偶尔的烤蛇为食,等待毛泽东的激进主义消退。当中国的一小群半导体工程师在中国的田野里锄地时,毛主义者却在劝告中国那些“所有人都必须制造半导体”,好像中国无产阶级的每个成员都可以在家中制造芯片。

In the 1960s, Mao won the political struggle over the Chinese semiconductor industry, downplaying its importance and cutting its ties with foreign technology. Most of China’s scientists resented the chairman for ruining their research—and their lives—by sending them to live on peasant farms to study proletarian politics rather than semiconductor engineering. One leading Chinese expert in optics who was sent to the countryside survived rural reeducation on a diet of rough grains, boiled cabbage, and an occasional grilled snake, as he waited for Mao’s radicalism to subside. While China’s small cadre of semiconductor engineers were hoeing China’s fields, Maoists exhorted the country’s workers that “all people must make semiconductors,” as if every member of the Chinese proletariat could forge chips at home.

中国领土的一小部分,躲过了文化大革命的恐怖。由于殖民主义的怪癖,香港仍然由英国人暂时统治。由于大多数中国人都在认真地记住他们疯狂的董事长的话,香港工人正在飞兆公司俯瞰九龙湾的工厂勤奋地组装硅元件。在几百英里外的台湾,多家美国芯片公司拥有雇佣数千名工人的设施,这些工人的工资按照加州的标准来说是低薪的,但远比农民农业好得多。就在毛泽东把中国的一小部分技术工人送到农村接受社会主义再教育时,台湾、韩国和整个东南亚的芯片产业正在把农民从农村拉出来,让他们在制造厂找到好工作。

One tiny speck of Chinese territory escaped the horrors of the Cultural Revolution. Thanks to a quirk of colonialism, Hong Kong was still governed temporarily by the British. As most Chinese were meticulously memorizing the quotations of their crazed chairman, Hong Kong workers were diligently assembling silicon components at Fairchild’s plant overlooking Kowloon Bay. A couple hundred miles away in Taiwan, multiple U.S. chip firms had facilities employing thousands of workers in jobs that were low-paying by California’s standards but far better than peasant farming. Just as Mao was sending China’s small set of skilled workers to the countryside for socialist reeducation, the chip industry in Taiwan, South Korea, and across Southeast Asia was pulling peasants from the countryside and giving them good jobs at manufacturing plants.

1970 年代初,随着毛的健康状况下降,文化大革命开始减弱。共产党领导人最终将科学家从农村召回。他们试着在实验室里捡拾碎片。但中国的芯片产业,在文革前远远落后于硅谷,现在也远远落后于中国的邻国。在中国陷入革命性混乱的十年中,英特尔发明了微处理器,而日本则在全球 DRAM 市场上占据了很大份额。除了骚扰其最聪明的公民外,中国一事无成。因此,到 1970 年代中期,其芯片产业处于灾难性的状态。“在我们生产的每 1,000 个半导体中,只有一个符合标准,”一位党的领导人在 1975 年抱怨道。“浪费了这么多。”

The Cultural Revolution began to wane as Mao’s health declined in the early 1970s. Communist Party leaders eventually called scientists back from the countryside. They tried picking up the pieces in their labs. But China’s chip industry, which had lagged far behind Silicon Valley before the Cultural Revolution, was now far behind China’s neighbors, too. During the decade in which China had descended into revolutionary chaos, Intel had invented microprocessors, while Japan had grabbed a large share of the global DRAM market. China accomplished nothing beyond harassing its smartest citizens. By the mid-1970s, therefore, its chip industry was in a disastrous state. “Out of every 1,000 semiconductors we produce, only one is up to standard,” one party leader complained in 1975. “So much is being wasted.”

1975 年 9 月 2 日,约翰·巴丁 (John Bardeen) 抵达北京,这是在他与 Shockley 和 Brattain 因发明晶体管而获得第一个诺贝尔奖的 20 年后。1972 年,他成为唯一一位第二次获得诺贝尔物理学奖的人,这一次是因为他在超导方面的工作。在物理学界,没有人比他更出名,尽管 Bardeen 是一个谦虚的人,他在 1940 年代后期被 Shockley 不公平地超越了。随着他接近退休,他将更多的时间投入到建立美国和外国大学之间的联系。1975 年,当一个由美国著名物理学家组成的代表团集结访问中国时,巴丁被邀请加入。

On September 2, 1975, John Bardeen landed in Beijing, two decades after he’d won his first Nobel Prize with Shockley and Brattain for inventing the transistor. In 1972, he had become the only person to win a second Nobel in physics, this time for work on superconductivity. In the world of physics, no one was more renowned, though Bardeen was the same modest man who’d been unfairly outshone by Shockley in the late 1940s. As he approached retirement, he devoted more time to building connections between American and foreign universities. When a delegation of prominent American physicists was being assembled to visit China in 1975, Bardeen was asked to join.

随着文化大革命的结束,中国领导人试图放下他们的革命热情,与美国人交好。巴丁来访时,毛病了。他会在第二年死去。巴丁的代表团提醒中国人与美国的友谊可以提供的技术。这次访问标志着自文化大革命以来发生了多么大的变化。十年前,诺贝尔奖获得者会被谴责为反革命代理人,不会受到北京、上海、南京和西安的中国领先研究机构的欢迎。但是,毛主义的大部分遗产仍然存在。美国人被告知,中国科学家没有发表他们的研究,因为他们反对“自夸。”

With the Cultural Revolution winding down, China’s leaders were trying to set aside their revolutionary fervor and befriend the Americans. At the time of Bardeen’s visit, Mao was ill; he would die the next year. Bardeen’s delegation reminded the Chinese of the technology that friendship with America could provide. This visit was a sign of how much had changed since the depths of the Cultural Revolution. A decade earlier, the Nobel Prize winner would have been denounced as a counterrevolutionary agent and not welcomed by China’s leading research institutes in Beijing, Shanghai, Nanjing, and Xian. But still, much of the Maoist legacy remained. The Americans were told that Chinese scientists didn’t publish their research because they opposed “self-glorification.”

巴丁从与肖克利的合作中了解了一些痴迷于自我美化的科学家,肖克利不公平地声称发明了晶体管的所有功劳。肖克利的例子——一位才华横溢的科学家,但一个失败的商人——证明了资本主义与自我美化之间的联系并不像毛泽东主义所暗示的那样简单。巴丁告诉他的妻子,尽管声称平等,但他发现中国社会受到控制和等级森严。这监视中国半导体科学家的政治头脑在硅谷当然没有可比性。

Bardeen knew something about scientists obsessed with self-glorification from his work with Shockley, who unfairly claimed all the credit for inventing the transistor. The example of Shockley—a brilliant scientist but a failed businessman—demonstrated that the link between capitalism and self-glorification wasn’t as straightforward as Maoist doctrine suggested. Bardeen told his wife that despite claims of equality he found Chinese society regimented and hierarchical. The political minders who watched over China’s semiconductor scientists certainly had no parallel in Silicon Valley.

Bardeen 和他的同事离开中国时对中国的科学家印象深刻,但中国的半导体制造野心似乎没有希望了。亚洲的电子革命已经完全从中国大陆过去了。硅谷芯片公司在从香港到台湾,从槟城到新加坡的工厂雇佣了数千名工人,通常是华裔。但中华人民共和国在 1960 年代一直在谴责资本家,而其邻国则拼命试图吸引他们。1979 年的一项研究发现,中国几乎没有任何商业上可行的半导体生产,只有全国一千五百台电脑。

Bardeen and his colleagues left China impressed with the country’s scientists, but China’s semiconductor manufacturing ambitions seemed hopeless. Asia’s electronics revolution had completely passed by mainland China. Silicon Valley chip firms employed thousands of workers, often ethnic Chinese, in plants from Hong Kong to Taiwan, Penang to Singapore. But the People’s Republic had spent the 1960s denouncing capitalists while its neighbors were trying desperately to attract them. A study in 1979 found that China had hardly any commercially viable semiconductor production and only fifteen hundred computers in the entire country.

毛泽东在巴丁访华后的第二年去世。几年后,老独裁者被邓小平取代,邓小平承诺实施“四个现代化”的政策来改造中国。不久,中国政府宣布“科学技术”是“四个现代化的症结”。世界其他地区正在被一场技术革命所改变,而中国的科学家们意识到芯片是这场变革的核心。1978 年 3 月召开的全国科学大会,正值邓小平巩固政权之际,将半导体置于议程的中心,希望中国能利用半导体的进步,帮助发展新的武器系统、消费电子产品和计算机。

Mao Zedong died the year after Bardeen’s visit to China. The old dictator was replaced, after a few years, by Deng Xiaoping, who promised a policy of “Four Modernizations” to transform China. Soon China’s government declared that “science and technology” were “the crux of the Four Modernizations.” The rest of the world was being transformed by a technological revolution, and China’s scientists realized that chips were at the core of this change. The National Science Conference held in March 1978, just as Deng Xiaoping was consolidating power, placed semiconductors at the center of its agenda, hoping that China could use advances in semiconductors to help develop new weapons systems, consumer electronics, and computers.

政治目标很明确:中国需要自己的半导体,不能依赖外国人。报纸《光明日报》定调,1985年号召读者摒弃“‘进口一机二机三机’的公式”,代之以“进口一机二造”。在中国,和第三台机器出口。” 这种对“中国制造”的痴迷已经根植于共产党的世界观中,但中国在半导体技术方面却无可救药地落后了——无论是毛泽东的大规模动员还是邓小平的命令都无法轻易改变。

The political goal was clear: China needed its own semiconductors, and it couldn’t rely on foreigners. Newspaper Guangming Ribao set the tone, calling on readers in 1985 to abandon “the formula of ‘the first machine imported, the second machine imported, and the third machine imported’ ” and replace it with “ ‘the first machine imported, the second made in China, and the third machine exported.’ ” This “Made in China” obsession was hardwired into the Communist Party’s worldview, but the country was hopelessly behind in semiconductor technology—something that neither Mao’s mass mobilization nor Deng’s diktat could easily change.

北京呼吁进行更多的半导体研究,但仅靠政府法令无法产生科学发明或可行的产业。政府坚持认为芯片具有重要的战略意义,导致中国官员试图控制芯片制造,使该行业陷入官僚主义。当像华为的任正非这样的新兴企业家在 1980 年代后期开始建立电子业务时,他们别无选择,只能依赖外国芯片。中国的电子组装产业建立在外国硅的基础上,从美国、日本和越来越多的台湾进口——共产党仍然认为台湾是“中国”的一部分,但仍不受其控制。

Beijing called for more semiconductor research, but government decrees alone couldn’t produce scientific inventions or viable industries. The government’s insistence that chips were strategically important caused China’s officials to try to control chipmaking, embroiling the sector in bureaucracy. When rising entrepreneurs like Huawei’s Ren Zhengfei began building electronics businesses in the late 1980s, they had no choice but to rely on foreign chips. China’s electronics assembly industry was built on a foundation of foreign silicon, imported from the United States, Japan, and increasingly Taiwan—which the Communist Party still considered part of “China,” but which remained outside its control.

第 31 章 “与中国人分享上帝的爱”

CHAPTER 31 “Sharing God’s Love with the Chinese”

理查德·张只是想“与中国人分享上帝的爱。” 圣经并没有太多关于半导体的内容,但张有传教士般的热情,将先进的芯片制造技术带到中国。一位虔诚的基督徒,出生于南京,在台湾长大,在德克萨斯州接受过培训的半导体工程师,他在 2000 年说服北京的统治者给他巨额补贴,以便在上海建立一家半导体代工厂。该设施完全按照他的规格设计,即使包括一座教堂,这要归功于中国通常是无神论者的政府的特别许可。如果张最终能够为他们带来现代半导体制造,该国领导人愿意在他们反对宗教的问题上妥协。然而,即使在政府的全力支持下,在与半导体行业的巨头,尤其是台湾的台积电斗争时,Chang 仍然觉得自己像大卫一样。

Richard Chang just wanted to “share God’s love with the Chinese.” The Bible didn’t say much about semiconductors, but Chang had a missionary’s zeal to bring advanced chipmaking to China. A devout Christian, the Nanjing-born, Taiwan-raised, Texas-trained semiconductor engineer convinced Beijing’s rulers in 2000 to give him vast subsidies to build a semiconductor foundry in Shanghai. The facility was designed exactly to his specifications, even including a church, thanks to special permission from China’s normally atheist government. The country’s leaders were willing to compromise on their opposition to religion if Chang could finally bring them modern semiconductor fabrication. Yet even with the full-fledged support of the government, Chang still felt like David as he struggled with the semiconductor industry’s goliaths, especially Taiwan’s TSMC.

芯片制造的地域在 1990 年代和 2000 年代发生了巨大变化。1990 年美国晶圆厂生产了全球 37% 的芯片,但到 2000 年这一数字下降到 19%,到 2010 年下降 13%。日本在芯片制造领域的市场份额也崩溃了。韩国、新加坡和台湾都纷纷向芯片产业注入资金,产量迅速增加。例如,新加坡政府与德州仪器、惠普和日立等公司合作,资助制造设施和芯片设计中心,在这个城市国家建立一个充满活力的半导体行业。新加坡政府也尝试复制台积电,建立代工厂,名为特许半导体,尽管该公司从未像其台湾竞争对手那样表现出色。

The geography of chip fabrication shifted drastically over the 1990s and 2000s. U.S. fabs made 37 percent of the world’s chips in 1990, but this number fell to 19 percent by 2000 and 13 percent by 2010. Japan’s market share in chip fabrication collapsed, too. South Korea, Singapore, and Taiwan each poured funds into their chip industries and rapidly increased output. For example, Singapore’s government funded fabrication facilities and chip design centers in partnership with companies like Texas Instruments, Hewlett-Packard, and Hitachi, building a vibrant semiconductor sector in the city-state. The Singaporean government also tried replicating TSMC, establishing a foundry called Chartered Semiconductor, though the company never performed as well as its Taiwanese rival.

韩国的半导体产业做得更好。在 1992 年取代日本的 DRAM 生产商并成为世界领先的内存芯片制造商之后,三星在接下来的十年中迅速成长。它在 DRAM 市场抵御来自台湾和新加坡的竞争,受益于政府的正式支持和政府对韩国银行提供信贷的非官方压力。这笔融资很重要,因为三星的主要产品 DRAM 内存芯片需要强大的财务力量才能到达每个连续的技术节点——即使在行业低迷时期也必须维持这种支出。DRAM市场是一位三星高管解释说,这就像一场小鸡游戏。景气时,全球DRAM企业纷纷将资金投入新工厂,将市场推向产能过剩,压低价格。继续支出是非常昂贵的,但停止投资,即使是一年,也有可能将市场份额拱手让给竞争对手。没有人想先眨眼。在竞争对手被迫削减开支后,三星有资金继续投资。它的存储芯片市场份额无情地增长。

South Korea’s semiconductor industry did even better. After dethroning Japan’s DRAM producers and becoming the world’s leading memory chipmaker in 1992, Samsung grew rapidly through the rest of that decade. It fended off competition in the DRAM market from Taiwan and Singapore, benefitting from formal government support and from unofficial government pressure on South Korea’s banks to provide credit. This financing mattered because Samsung’s main product, DRAM memory chips, required brute financial force to reach each successive technology node—spending that had to be sustained even during industry downturns. The DRAM market was like a game of chicken, one Samsung executive explained. In good times, the world’s DRAM companies would pour money into new factories, pushing the market toward overcapacity, driving down prices. Carrying on spending was ruinously expensive, but stopping investments, even for a single year, risked ceding market share to rivals. No one wanted to blink first. Samsung had the capital to keep investing after its rivals were forced to cut back. Its memory chip market share grew inexorably.

中国最有可能颠覆半导体行业,因为它在组装世界上大多数芯片的电子设备方面发挥着越来越大的作用。到 1990 年代,自从该国第一次不幸的半导体生产努力被毛派激进主义打断以来,已经过去了几十年。中国已成为世界工厂,上海和深圳等城市是电子组装中心——几十年前推动台湾经济的工作类型。然而,中国领导人知道真正的钱是在驱动电子产品的组件上,尤其是在半导体上。

China had the most potential to upend the semiconductor industry, given its growing role assembling the electronic devices into which most of the world’s chips were slotted. By the 1990s, decades had passed since the country’s first ill-fated efforts at semiconductor production were interrupted by Maoist radicalism. China had become the world’s workshop, and cities like Shanghai and Shenzhen were centers of electronics assembly—the type of work that had propelled Taiwan’s economy several decades earlier. However, China’s leaders knew the real money was in the components that powered electronics, above all in semiconductors.

1990年代中国的芯片制造能力远远落后于台湾和韩国,更不用说美国了。尽管中国的经济改革如火如荼,走私者仍然发现通过以下方式非法将芯片带入该国是有利可图的塞满行李箱,从香港过境。但随着中国电子产业的成熟,走私芯片开始变得没有制造芯片那么有吸引力。

China’s chip manufacturing capabilities in the 1990s lagged far behind Taiwan and South Korea, to say nothing of the United States. Even though China’s economic reforms were in full swing, smugglers still found it profitable to bring chips illegally into the country by stuffing suitcases full of them and crossing the border from Hong Kong. But as China’s electronics industry matured, smuggling chips began to seem less appealing than making them.

Richard Chang 认为将芯片带到中国是他一生的使命。1948年出生于前首都南京的一个军人家庭,他的家人在共产党掌权后逃离中国,年仅一岁时抵达台湾。在台湾,他在一个大陆人社区长大,他们将在岛上居住视为临时逗留。人民共和国的预期崩溃从未到来,让像张这样的人处于永久的身份危机状态,将自己视为中国人,但生活在一个从政治角度讲,离他们出生的土地越来越远的岛上。大学毕业后,Chang 移居美国,在纽约布法罗完成了研究生学位,之后在德州仪器 (Texas Instruments) 工作,与杰克·基尔比 (Jack Kilby) 一起工作。他成为了运营晶圆厂的专家,从美国到日本,从新加坡到意大利,TI 的设施遍布全球。

Richard Chang saw bringing chips to China as his life’s calling. Born in 1948 to a military family in Nanjing, the former capital, his family fled China after the Communists took power, arriving in Taiwan when he was only one year old. In Taiwan, he grew up in a community of mainlanders who treated residence on the island as a temporary sojourn. The expected collapse of the People’s Republic never came, leaving people like Chang in a permanent state of identity crisis, seeing themselves as Chinese but living on an island that, in political terms, was drifting ever further away from the land of their birth. After finishing university, Chang moved to the U.S., completing a graduate degree in Buffalo, New York, before taking a job at Texas Instruments, where he worked with Jack Kilby. He became an expert in operating fabs, running TI’s facilities around the world, from the U.S. to Japan, Singapore to Italy.

中国政府补贴国内半导体产业建设的大部分早期成果并不令人印象深刻。一些晶圆厂是在中国建造的,例如中国华虹和日本NEC在上海的合资企业。NEC收到了来自中国政府的甜蜜金融交易,以换取承诺将其技术带到中国。但是,NEC 确保由日本专家负责。中国工人只被允许从事基本活动。“我们不能说这个行业是中国的行业,”一位分析师被引述说。这只是“位于中国的晶圆厂”。中国从合资企业中获得的专业知识很少。

Most of the early results of China’s government efforts to subsidize the construction of a domestic semiconductor industry weren’t impressive. Some fabs were built in China, such as a joint venture in Shanghai between China’s Huahong and Japan’s NEC. NEC received a sweet financial deal from the Chinese government in exchange for promising to bring its technology to China. However, NEC made sure that Japanese experts were in charge; Chinese workers were only allowed to undertake basic activities. “We cannot say this industry is a Chinese industry,” one analyst was quoted as saying. It was just a “wafer fab located in China.” China gained little expertise from the joint venture.

另一家于 2000 年在上海成立的芯片公司格雷斯半导体也涉及类似的外国投资、国家补贴和技术转让失败。格蕾丝是江之间的冒险中国国家主席江泽民之子绵恒与一国之子王文斯顿台湾塑料王朝。鉴于台湾在半导体领域的成功,吸引台湾人参与中国芯片产业的想法是有道理的,而中国国家主席的孩子的参与有助于获得政府的支持。公司甚至聘请了乔治·W·布什总统的弟弟尼尔·布什为“商业战略”提供建议,每年向他支付 40 万美元以表彰他的洞察力。这个星光熠熠的领导团队可能让格蕾丝远离了政治麻烦,但公司的技术落后了努力争取客户,在中国代工业务中的份额从未超过世界份额的一小部分。

Grace Semiconductor, another chip firm founded in Shanghai, in 2000, involved a similar mix of foreign investment, state subsidies, and failed technology transfer. Grace was a venture between Jiang Mianheng, son of Chinese president Jiang Zemin, and Winston Wang, scion of a Taiwanese plastics dynasty. The idea of attracting Taiwanese participation in China’s chip industry made sense given the island’s success in semiconductors, while the involvement of a child of a Chinese president helped secure government support. The company even hired Neil Bush, a younger brother of President George W. Bush, to advise on “business strategies,” paying him $400,000 annually for his insight. This star-studded leadership team may have kept Grace out of political trouble, but the company’s technology lagged and it struggled to acquire customers, never winning more than a small share of China’s foundry business, a sliver of the world’s total.

如果有人能在中国建立一个芯片产业,那就是 Richard Chang。他不会依赖裙带关系或外国帮助。世界级晶圆厂所需的所有知识都已经在他的脑海中。在德州仪器工作期间,他为该公司在世界各地开设了新设施。为什么他不能在上海做同样的事情?他于 2000 年创立了中芯国际(SMIC),从国际投资者那里筹集了超过 15 亿美元的资金,例如高盛、摩托罗拉和东芝。一位分析师估计,中芯国际一半的启动资金由美国投资者提供。Chang利用这些资金聘请了数百名外国人来经营中芯国际的晶圆厂,其中至少包括台湾四百。

If anyone could build a chip industry in China, it was Richard Chang. He wouldn’t rely on nepotism or on foreign help. All the knowledge needed for a world-class fab was already in his head. While working at Texas Instruments, he’d opened new facilities for the company around the world. Why couldn’t he do the same in Shanghai? He founded the Semiconductor Manufacturing International Corporation (SMIC) in 2000, raising over $1.5 billion from international investors like Goldman Sachs, Motorola, and Toshiba. One analyst estimated that half of SMIC’s startup capital was provided by U.S. investors. Chang used these funds to hire hundreds of foreigners to operate SMIC’s fab, including at least four hundred from Taiwan.

张的策略很简单:像台积电那样做。在台湾,台积电聘请了它所能找到的最好的工程师,最好是在美国或其他先进芯片公司工作过的工程师。台积电购买了它买得起的最好的工具。它坚持不懈地专注于对员工进行行业最佳实践的培训。它利用了台湾政府愿意提供的所有税收和补贴。

Chang’s strategy was simple: do as TSMC had done. In Taiwan, TSMC had hired the best engineers it could find, ideally with experience at American or other advanced chip firms. TSMC bought the best tools it could afford. It focused relentlessly on training its employees in the industry’s best practices. And it took advantage of all the tax and subsidy benefits that Taiwan’s government was willing to provide.

中芯国际虔诚地遵循了这个路线图。它从海外芯片制造商那里大举招聘,尤其是台湾。在中芯国际运营的第一个十年的大部分时间里,中芯国际三分之一的工程人员都是从海外招聘的。据分析师 Doug Fuller 称,2001 年,中芯国际雇佣了 650 名本地工程师,而从海外招聘的工程师为 393 人,大部分来自台湾和美国 到本世纪末,大约三分之一的工程员工是从国外招聘的。公司甚至有一个口号,“一个老员工带两个新员工”,强调需要有经验的外国培训员工来帮助当地工程师学习。中芯国际的本地工程师学得很快,很快就被认为能力很强,他们开始收到外国芯片制造商的工作机会。公司在驯化技术方面的成功,也正是得益于此受过外国培训的劳动力。

SMIC followed this road map religiously. It hired aggressively from overseas chipmakers, especially from Taiwan. For much of its first decade of operation, a third of SMIC’s engineering personnel were hired from overseas. In 2001, according to analyst Doug Fuller, SMIC employed 650 local engineers compared with 393 who were recruited from overseas, mostly from Taiwan and the U.S. Through the end of the decade, roughly a third of engineering employees were hired from abroad. The company even had a slogan, “one old staffer brings along two new ones,” emphasizing the need for experienced foreign-trained employees to help local engineers learn. SMIC’s local engineers learned quickly, and were soon perceived to be so capable they began receiving job offers from foreign chipmakers. The company’s success in domesticating technology was only possible thanks to this foreign-trained workforce.

与中国其他芯片初创公司一样,中芯国际受益于政府的大力支持,例如五年的企业免税期和降低在中国销售的芯片的销售税。中芯国际榨取了这些好处,但起初它并不依赖于它们。与那些更注重雇佣政客子女而不是制造质量的竞争对手不同,Chang 提高了生产能力并采用了靠近尖端。到 2000 年代末,中芯国际仅落后于世界技术领先者几年。该公司似乎有望成为一流的代工厂,或许最终有能力威胁台积电。Richard Chang 很快赢得了为他的前雇主德州仪器等行业领导者制造芯片的合同。中芯国际于 2004 年在纽约证券交易所上市。

Like China’s other chip startups, SMIC benefitted from vast government support, like a five-year corporate tax holiday and reduced sales tax on chips sold in China. SMIC milked these benefits, but at first it didn’t depend on them. Unlike rivals who focused more on hiring politicians’ children than on manufacturing quality, Chang ramped up production capacity and adopted technology that was near the cutting edge. By the end of the 2000s SMIC was only a couple years behind the world’s technology leaders. The company seemed on track to become a top-notch foundry, perhaps eventually capable of threatening TSMC. Richard Chang soon won contracts to build chips for industry leaders like his former employer, Texas Instruments. SMIC listed its shares on the New York Stock Exchange in 2004.

现在台积电面临来自东亚不同国家的多家代工厂的竞争。新加坡的特许半导体、台湾的联电和先锋半导体以及 2005 年进入代工业务的韩国三星也在与台积电竞争生产其他地方设计的芯片。这些公司中的大多数都得到了政府的补贴,但这使得芯片生产更便宜,使他们所服务的大多数美国无晶圆厂半导体设计师受益。与此同时,无晶圆厂公司正处于推出充满复杂芯片的革命性新产品的早期阶段:智能手机。离岸外包降低了制造成本并刺激了更多竞争。消费者受益于低廉的价格和以前无法想象的设备。这不正是全球化的设计目的吗?

Now TSMC had competition from multiple foundries in different countries in East Asia. Singapore’s Chartered Semiconductor, Taiwan’s UMC and Vanguard Semiconductor, and South Korea’s Samsung—which entered the foundry business in 2005—were also competing with TSMC to produce chips designed elsewhere. Most of these companies were subsidized by their governments, but this made chip production cheaper, benefitting the mostly American fabless semiconductor designers they served. Fabless firms, meanwhile, were in the early stages of launching a revolutionary new product chock-full of complex chips: the smartphone. Offshoring had reduced manufacturing costs and spurred more competition. Consumers benefitted from low prices and from previously unthinkable devices. Wasn’t this exactly how globalization was designed to work?

第 32 章 光刻战争

CHAPTER 32 Lithography Wars

1992 年,当 John Carruthers在位于加利福尼亚州圣克拉拉的英特尔总部的会议室里坐下时,他没想到向英特尔 CEO 安迪·格鲁夫 (Andy Grove) 索要 2 亿美元会很容易。作为英特尔研发工作的领导者,Carruthers 习惯于下大赌注。有些工作,有些则没有,但英特尔工程师的击球率与业内任何人一样好。到 1992 年,英特尔再次成为世界上最大的芯片制造商,这得益于 Grove 决定将英特尔的精力集中在 PC 微处理器上。它现金充裕,并且一如既往地致力于摩尔定律。

When John Carruthers sat down in a meeting room at Intel’s headquarters in Santa Clara, California, in 1992, he didn’t expect that asking Intel CEO Andy Grove for $200 million was going to be easy. As a leader of Intel’s R&D efforts, Carruthers was used to making big bets. Some worked, and others didn’t, but Intel’s engineers had as good a batting average as anyone in the industry. By 1992, Intel was again the world’s biggest chipmaker, on the strength of Grove’s decision to focus Intel’s efforts on microprocessors for PCs. It was flush with cash and as committed as ever to Moore’s Law.

然而,Carruthers 的要求远远超出了通常的研发项目。与业内其他人一样,Carruthers 知道现有的光刻方法很快将无法生产下一代半导体所需的更小电路。光刻公司正在推出使用深紫外光的工具,波长为 248 或 193 纳米,人眼不可见。但不久之后,芯片制造商就会要求更高的光刻精度。他想瞄准波长为 13.5 纳米的“极紫外”(EUV)光。波长越小,可以刻在芯片上的特征就越小。只有一个问题:大多数人认为极紫外光不可能大规模生产。

Carruthers’s request stretched far beyond the usual for R&D projects, however. Along with everyone else in the industry, Carruthers knew existing lithography methods would soon be unable to produce the ever-smaller circuits that next-generation semiconductors required. Lithography companies were rolling out tools using deep ultraviolet light, with wavelengths of 248 or 193 nanometers, invisible to the human eye. But it wouldn’t be long before chipmakers would be asking for even more lithographic precision. He wanted to target “extreme ultraviolet” (EUV) light, with a wavelength of 13.5 nanometers. The smaller the wavelength, the smaller the features that could be carved onto chips. There was only one problem: most people thought extreme ultraviolet light was impossible to mass-produce.

“你的意思是告诉我你要花钱买一些我们甚至不知道它是否会起作用的东西?” 格鲁夫怀疑地问道。“是的,安迪,这就是所谓的研究,”卡拉瑟斯反驳道。格鲁夫求助于英特尔前首席执行官戈登·摩尔,后者仍然是该公司的顾问。“你会怎么做,戈登?” “好吧,安迪,你还有什么选择?” 摩尔问道。答案很明显:没有。芯片行业要么学会使用更小的波长进行光刻,要么晶体管的缩小——以及以摩尔命名的定律——将停止。这样的结果对英特尔的业务将是毁灭性的,对格鲁夫来说将是耻辱。他给了 Carruthers 2 亿美元用于开发 EUV 光刻技术。英特尔最终将花费数十亿美元用于研发,并花费数十亿美元学习如何使用 EUV 雕刻芯片。它从未计划制造自己的 EUV 设备,但需要保证至少有一家世界先进的光刻公司将 EUV 机器推向市场,以便英特尔拥有雕刻更小的电路所需的工具。

“You mean to tell me you’re going to spend money on something that we don’t even know if it’s gonna work?” Grove asked skeptically. “Yeah, Andy, that’s called research,” Carruthers retorted. Grove turned to Gordon Moore, Intel’s former CEO, who remained an advisor to the company. “What would you do, Gordon?” “Well, Andy, what other choices do you have?” Moore asked. The answer was obvious: none. The chip industry would either learn to use ever smaller wavelengths for lithography, or the shrinking of transistors—and the law named after Moore—would come to a halt. Such an outcome would be devastating for Intel’s business and humiliating for Grove. He gave Carruthers $200 million to spend developing EUV lithography. Intel would eventually spend billions of dollars on R&D and billions more learning how to use EUV to carve chips. It never planned to make its own EUV equipment, but needed to guarantee that at least one of the world’s advanced lithography firms would bring EUV machines to market so that Intel would have the tools needed to carve ever-smaller circuits.

自从 Jay Lathrop 在他的美国军事实验室将他的显微镜倒置以来,在 1990 年代,光刻技术的未来受到质疑。光刻行业面临三个生存问题:工程、商业和地缘政治。在芯片制造的早期,晶体管是如此之大,以至于光刻工具使用的光波大小几乎无关紧要。但是摩尔定律已经发展到光波的尺度——几百纳米,取决于颜色——会影响蚀刻电路的精度。到 1990 年代,最先进的晶体管的测量尺寸为数百纳米(十亿分之一米),但已经可以设想长度只有十几纳米的更小晶体管。

More than at any point since Jay Lathrop had turned his microscope upside down in his U.S. military lab, in the 1990s the future of lithography was in doubt. Three existential questions hung over the lithography industry: engineering, business, and geopolitics. In the early days of chipmaking, transistors were so big that the size of the light waves used by lithography tools barely mattered. But Moore’s Law had progressed to the point where the scale of light waves—a couple hundred nanometers, depending on the color—impacted the precision with which circuits could be etched. By the 1990s, the most advanced transistors were measured in the hundreds of nanometers (billionths of a meter), but it was already possible to envision far smaller transistors with features just a dozen nanometers in length.

大多数研究人员认为,生产这种规模的芯片需要更精确的光刻工具来照射光刻胶化学品并在硅上雕刻形状。一些研究人员试图使用电子束来雕刻芯片,但电子束光刻的速度从来都不足以进行大规模生产。其他人则将赌注押在 X 光片上或极紫外光,每一种都与不同的光刻胶化学物质发生反应。在一年一度的国际光刻专家会议上,科学家们讨论了哪种技术会胜出。那是一个时代一位参与者说,“光刻战争”是工程师之间相互竞争的群体。

Producing chips at this scale, most researchers believed, required more precise lithography tools to shoot light at photoresist chemicals and carve shapes on silicon. Some researchers sought to use beams of electrons to carve chips, but electron beam lithography was never fast enough for mass production. Others placed their bet on X-rays or extreme ultraviolet light, each of which reacted with different sets of photoresist chemicals. At the annual international conference of lithography experts, scientists debated which technique would win out. It was a time of “lithography wars,” one participant put it, between competing groups of engineers.

寻找下一个最好的射束类型以射向硅晶片的“战争”只是光刻技术未来正在进行的三场竞赛之一。第二场战斗是商业上的,关于哪家公司将制造下一代光刻工具。开发新光刻设备的巨额成本推动了该行业的集中化。一家或至多两家公司将主导市场。在美国,GCA 已被清算,而 Perkin Elmer 的后裔光刻公司硅谷集团则远远落后于市场领导者佳能和尼康。美国芯片制造商在 1980 年代抵御了日本的挑战,但美国的光刻工具制造商却没有。

The “war” to find the next, best type of beam to shoot at silicon wafers was only one of three contests underway over the future of lithography. The second battle was commercial, over which company would build the next generation of lithography tools. The enormous cost of developing new lithography equipment pushed the industry toward concentration. One or at most two companies would dominate the market. In the United States, GCA had been liquidated, while Silicon Valley Group, a lithography firm descended from Perkin Elmer, lagged far behind the market leaders, Canon and Nikon. U.S. chipmakers had fended off the Japanese challenge of the 1980s, but American lithography toolmakers hadn’t.

佳能和尼康唯一真正的竞争对手是 ASML,这是一家规模虽小但正在成长的荷兰光刻公司。1984 年,荷兰电子公司飞利浦剥离了其内部的光刻部门,创建了 ASML。恰逢芯片价格暴跌导致 GCA 业务下滑,分拆的时机非常糟糕。更重要的是,距离荷兰和比利时边境不远的小镇维尔德霍芬似乎不太可能成为半导体行业的世界级公司。欧洲是一个相当大的芯片生产国,但它明显落后于硅谷和日本。

The only real competitor to Canon and Nikon was ASML, the small but growing Dutch lithography company. In 1984, Philips, the Dutch electronics firm, had spun out its internal lithography division, creating ASML. Coinciding with the collapse in chip prices that sank GCA’s business, the spinoff was horribly timed. What’s more, Veldhoven, a town not far from the Dutch border with Belgium, seemed an unlikely place for a world-class company in the semiconductor industry. Europe was a sizeable producer of chips, but it was very clearly behind Silicon Valley and Japan.

1984 年,荷兰工程师 Frits van Hout 在完成物理学硕士学位后加入 ASML,该公司的员工问他是否愿意自愿加入或被迫接受这份工作。除了与飞利浦的合作之外,“我们有没有设施,也没有钱,”范豪特回忆道。为光刻工具建立庞大的内部制造工艺是不可能的。相反,该公司决定使用从世界各地的供应商那里精心采购的组件来组装系统。依靠其他公司获得关键组件带来了明显的风险,但 ASML 学会了管理它们。日本竞争对手试图在内部制造所有东西,而 ASML 可以购买市场上最好的组件。随着它开始专注于开发 EUV 工具,其集成来自不同来源的组件的能力成为其最大的优势。

When Dutch engineer Frits van Hout joined ASML in 1984 just after completing his master’s degree in physics, the company’s employees asked whether he’d joined voluntarily or was forced to take the job. Beyond its tie with Philips, “we had no facilities and no money,” van Hout remembered. Building vast in-house manufacturing processes for lithography tools would have been impossible. Instead, the company decided to assemble systems from components meticulously sourced from suppliers around the world. Relying on other companies for key components brought obvious risks, but ASML learned to manage them. Whereas Japanese competitors tried to build everything in-house, ASML could buy the best components on the market. As it began to focus on developing EUV tools, its ability to integrate components from different sources became its greatest strength.

出乎意料的是,ASML 的第二个优势是它在荷兰的位置。在 1980 年代和 1990 年代,该公司在日本和美国之间的贸易争端中被视为中立。美国公司将其视为尼康和佳能的可靠替代品。例如,美国 DRAM 初创公司 Micron 想要购买光刻工具时,它转向了 ASML,而不是依赖于两家主要的日本供应商之一,这两家供应商都与美光的 DRAM 有着深厚的联系日本的竞争对手。

ASML’s second strength, unexpectedly, was its location in the Netherlands. In the 1980s and 1990s, the company was seen as neutral in the trade disputes between Japan and the United States. U.S. firms treated it like a trustworthy alternative to Nikon and Canon. For example, when Micron, the American DRAM startup, wanted to buy lithography tools, it turned to ASML rather than relying on one of the two main Japanese suppliers, each of which had deep ties with Micron’s DRAM competitors in Japan.

ASML 从飞利浦分拆出来的历史也以令人惊讶的方式促成了与台湾台积电的深厚关系。飞利浦一直是台积电的基石投资者,将其制造工艺技术和知识产权转让给了这家年轻的代工厂。这给了 ASML 一个内置市场,因为台积电的晶圆厂是围绕飞利浦的制造工艺设计的。1989 年台积电晶圆厂的一次意外火灾也起到了帮助作用,导致台积电购买了另外 19 台新的光刻机,由火灾保险支付。ASML 和 TSMC 都是从芯片行业外围的小公司开始的,但他们一起成长,没有这种伙伴关系,今天的计算进步就会停滞不前。

ASML’s history of being spun out of Philips helped in a surprising way, too, facilitating a deep relationship with Taiwan’s TSMC. Philips had been the cornerstone investor in TSMC, transferring its manufacturing process technology and intellectual property to the young foundry. This gave ASML a built-in market, because TSMC’s fabs were designed around Philips’s manufacturing processes. An accidental fire in TSMC’s fab in 1989 helped, too, causing TSMC to buy an additional nineteen new lithography machines, paid for by the fire insurance. Both ASML and TSMC started as small firms on the periphery of the chip industry, but they grew together, forming a partnership without which advances in computing today would have ground to a halt.

ASML 与台积电的合作指向了 1990 年代的第三次“光刻战争”。这是一场政治竞赛,尽管行业或政府中很少有人愿意以这种方式思考。当时,美国正在庆祝冷战结束并兑现其和平红利。以技术、军事或经济实力衡量,美国在世界其他地区、盟友和对手中都遥遥领先。一位有影响力的评论员宣称 1990 年代是“单极时刻”,美国的主导地位不容置疑。波斯湾战争展示了美国可怕的技术和军事实力。

The partnership between ASML and TSMC pointed to the third “lithography war” of the 1990s. This was a political contest, though few people in industry or government preferred to think in those terms. At the time, the U.S. was celebrating the end of the Cold War and cashing in its peace dividend. Measured by technological, military, or economic power, the U.S. towered above the rest of the world, allies and adversaries alike. One influential commentator declared the 1990s a “unipolar moment,” in which America’s dominance was unquestioned. The Persian Gulf War had demonstrated America’s terrifying technological and military might.

1992 年,当安迪·格鲁夫准备批准英特尔对 EUV 光刻研究的第一笔重大投资时,很容易理解为什么即使是从冷战军工联合体中崛起的芯片行业也得出结论认为政治不再重要。管理大师许诺未来“无国界世界”,其中利润而非权力将塑造全球商业格局。经济学家谈到加速全球化。CEO 和政治家都接受了这些新的知识时尚。与此同时,英特尔再次成为半导体业务的佼佼者。它已经抵御了日本的竞争对手,现在几乎垄断了个人电脑芯片的全球市场。它有自1986年以来每年都在盈利。它为什么要担心政治?

When Andy Grove was preparing to approve Intel’s first major investment in EUV lithography research in 1992, it was easy to see why even the chip industry, which had emerged out of the Cold War military-industrial complex, had concluded politics no longer mattered. Management gurus promised a future “borderless world” in which profits not power would shape the global business landscape. Economists spoke of accelerating globalization. CEOs and politicians alike embraced these new intellectual fashions. Intel, meanwhile, was again on top of the semiconductor business. It had fended off its Japanese rivals and now all but monopolized the global market for the chips that powered personal computers. It has made a profit every year since 1986. Why should it worry about politics?

1996 年,英特尔与美国能源部运营的几个实验室建立了合作伙伴关系,这些实验室在光学和其他领域的专业知识使 EUV 发挥作用。英特尔召集了六家其他芯片制造商加入该财团,但英特尔支付了大部分费用,并且是一位参与者记得,房间里有“95% 的大猩猩”。英特尔知道劳伦斯利弗莫尔和桑迪亚国家实验室的研究人员拥有构建 EUV 原型系统的专业知识,但他们的重点是科学,而不是大规模生产。

In 1996, Intel forged a partnership with several of the laboratories operated by the U.S. Department of Energy, which had expertise in optics and other fields needed to make EUV work. Intel assembled a half dozen other chipmakers to join the consortium, but Intel paid for most of it and was the “95 percent gorilla” in the room, one participant remembered. Intel knew that the researchers at Lawrence Livermore and Sandia National Labs had the expertise to build a prototype EUV system, but their focus was on the science, not on mass production.

英特尔的目标是“制造东西,而不仅仅是测量它”,Carruthers 解释说,因此该公司开始寻找一家将 EUV 工具商业化和大规模生产的公司。它得出的结论是,没有任何一家美国公司可以做到这一点。GCA 已不复存在。美国最大的光刻公司是硅谷集团(SVG),它在技术上落后。美国政府对 1980 年代的贸易战仍然敏感,不希望日本的尼康和佳能与国家实验室合作,尽管尼康本身并不认为 EUV 技术会起作用。ASML 是唯一剩下的光刻公司。

Intel’s goal was “to make stuff, not just to measure it,” Carruthers explained, so the company began searching for a company to commercialize and mass-produce EUV tools. It concluded no American firm could do it. GCA was no more. America’s biggest remaining lithography firm was Silicon Valley Group (SVG), which lagged technologically. The U.S. government, still sensitive from the trade wars of the 1980s, didn’t want Japan’s Nikon and Canon to work with the national labs, though Nikon itself didn’t think EUV technology would work. ASML was the only lithography firm left.

让外国公司获得来自美国国家实验室的最先进研究的想法提出了一些问题在华盛顿。EUV 技术没有立即的军事应用,目前还不清楚 EUV 是否会起作用。尽管如此,如果确实如此,美国将依赖 ASML 作为所有计算的基础工具。除了国防部的少数官员外,华盛顿几乎没有人担心。大多数人将 ASML 和荷兰政府视为可靠的合作伙伴。对政治领导人来说更重要的是影响工作,而不是地缘政治。美国政府要求 ASML 在美国建立工厂,为其光刻工具制造组件,并为美国客户提供产品并雇用美国员工。然而,ASML 的大部分核心研发将在荷兰进行。商务部、国家实验室和相关公司的主要决策者表示,他们不记得政治因素在政府决定让这一安排继续进行。

The idea of giving a foreign company access to the most advanced research coming out of America’s national labs raised some questions in Washington. There was no immediate military application for EUV technology, and it still wasn’t clear that EUV would work. Nevertheless, if it did, the U.S. would be reliant on ASML for a tool fundamental to all computing. Except for a few officials in the Defense Department, hardly anyone in Washington was concerned. Most people saw ASML and the Dutch government as reliable partners. More important to political leaders was the impact on jobs, not geopolitics. The U.S. government required ASML to build a facility in the U.S. to manufacture components for its lithography tools and supply American customers and employ American staff. However, much of ASML’s core R&D would take place in the Netherlands. Key decision makers from the Commerce Department, the National Labs, and the companies involved say they don’t recall political considerations playing much if any role in the government’s decision to let this arrangement proceed.

尽管延迟时间长且成本超支,但 EUV 合作伙伴关系缓慢取得了进展。由于美国国家实验室的研究被拒之门外,尼康和佳能决定不制造自己的 EUV 工具,让 ASML 成为世界上唯一的生产商。与此同时,2001 年,ASML 收购了美国最后一家大型光刻公司 SVG。SVG 已经远远落后于行业领导者,但人们再次质疑该交易是否适合美国的安全利益。几十年来一直资助光刻行业的 DARPA 和国防部内部,一些官员反对这笔交易。国会也提出了担忧,三名参议员给乔治·W·布什总统写信说:“ASML 最终将所有美国政府的 EUV 技术。”

Despite long delays and huge cost overruns, the EUV partnership slowly made progress. Locked out of the research at the U.S. national labs, Nikon and Canon decided not to build their own EUV tools, leaving ASML as the world’s only producer. In 2001, meanwhile, ASML bought SVG, America’s last major lithography firm. SVG already lagged far behind industry leaders, but again questions were raised about whether the deal suited America’s security interests. Inside DARPA and the Defense Department, which had funded the lithography industry for decades, some officials opposed the sale. Congress raised concerns, too, with three senators writing President George W. Bush that “ASML will wind up with all of the U.S. government’s EUV technology.”

不可否认,这是真的。但美国的实力正处于巅峰时期。华盛顿的大多数人认为全球化是一件好事。美国政府的主要信念是,扩大贸易和供应链联系将通过鼓励俄罗斯或中国等大国专注于获取财富而不是地缘政治力量来促进和平。声称美国光刻行业的衰落将危及安全的说法被视为与这个全球化和互联互通的新时代脱节。与此同时,芯片行业只是想尽可能高效地制造半导体。美国没有大型光刻机企业,除了押注ASML,他们还有什么选择?

This was undeniably true. But America’s power was at its peak. Most people in Washington thought globalization was a good thing. The dominant belief in the U.S. government was that expanding trade and supply chain connections would promote peace by encouraging powers like Russia or China to focus on acquiring wealth rather than geopolitical power. Claims that the decline of America’s lithography industry would imperil security were seen as out of touch with this new era of globalization and interconnection. The chip industry, meanwhile, simply wanted to build semiconductors as efficiently as possible. With no large-scale U.S. lithography firms remaining, what choice did they have but to bet on ASML?

英特尔和其他大型芯片制造商认为,将 SVG 出售给 ASML 对于开发 EUV 至关重要,因此对计算的未来至关重要。“如果没有合并,”英特尔新任首席执行官 Craig Barrett 在 2001 年指出,“新工具在美国的开发路径将被推迟。” 随着冷战的结束,刚刚掌权的布什政府希望放松对所有商品的技术出口管制,但直接用于军事用途的商品除外。政府将该战略描述为“围绕最敏感的技术建造高墙”。EUV 没有上榜。

Intel and other big chipmakers argued that the sale of SVG to ASML was crucial to developing EUV—and thus fundamental to the future of computing. “Without the merger,” Intel’s new CEO Craig Barrett argued in 2001, “the development path to the new tools in the U.S. will be delayed.” With the Cold War over, the Bush administration, which had just taken power, wanted to loosen technology export controls on all goods except those with direct military applications. The administration described the strategy as “building high walls around technologies of the highest sensitivity.” EUV didn’t make the list.

因此,下一代 EUV 光刻工具将主要在国外组装,尽管一些组件继续在康涅狄格州的工厂制造。任何提出美国如何保证获得 EUV 工具的问题的人都被指责在全球化的世界中保持冷战思维。然而,那些谈论技术在全球传播的商业大师歪曲了正在发挥作用的动态。产生 EUV 的科学网络遍布全球,汇集了来自不同国家的科学家美国、日本、斯洛文尼亚和希腊。但是,EUV 的制造并没有全球化,而是被垄断了。由一家公司管理的单一供应链将控制光刻的未来。

The next-generation EUV lithography tools would therefore be mostly assembled abroad, though some components continued to be built in a facility in Connecticut. Anyone who raised the question of how the U.S. could guarantee access to EUV tools was accused of retaining a Cold War mindset in a globalizing world. Yet the business gurus who spoke about technology spreading globally misrepresented the dynamic at play. The scientific networks that produced EUV spanned the world, bringing together scientists from countries as diverse as America, Japan, Slovenia, and Greece. However, the manufacturing of EUV wasn’t globalized, it was monopolized. A single supply chain managed by a single company would control the future of lithography.

第 33 章 创新者的困境

CHAPTER 33 The Innovator’s Dilemma

在 2006 年 Macworld 大会上,史蒂夫·乔布斯独自站在黑暗的舞台上,穿着他标志性的蓝色牛仔裤和黑色高领毛衣。数百名科技爱好者的听众焦急地等待着硅谷的先知发言。乔布斯向左转身,舞台远处冒出青烟。一个穿着白色兔子套装的男人——半导体工人用来保持工厂超清洁的那种——穿过烟雾,穿过舞台,一直走到乔布斯面前。他摘下头巾,咧嘴一笑:是英特尔 CEO 保罗·欧德宁。他递给乔布斯一块大硅片。“史蒂夫,我想报告英特尔准备好了。”

Steve Jobs stood alone on a dark stage at the 2006 Macworld conference, wearing his trademark blue jeans and a black turtleneck. An audience of hundreds of tech buffs waited anxiously for Silicon Valley’s prophet to speak. Jobs turned toward his left, and blue smoke erupted on the far side of the stage. A man in a white bunny suit—the type used by semiconductor workers to keep their fabs ultra-clean—walked through the smoke, across the stage, right up to Jobs. He took off his head covering and grinned: it was Intel CEO Paul Otellini. He handed Jobs a large silicon wafer. “Steve, I want to report that Intel is ready.”

这是经典的史蒂夫乔布斯剧院,但它是典型的英特尔商业政变。到 2006 年,英特尔已经为大多数 PC 提供了处理器,在过去十年中成功抵御了 AMD,这是唯一一家在 x86 指令集架构上生产芯片的主要公司——这是一套管理芯片计算方式的基本规则——那是个人电脑的行业标准。苹果是唯一一家不使用基于 x86 的芯片的主要计算机制造商。现在,乔布斯和欧德宁宣布,这将改变。Mac 计算机内部将装有英特尔芯片。英特尔的帝国将会壮大,它对 PC 行业的控制将会收紧。

This was classic Steve Jobs theater, but it was a typical Intel business coup. By 2006, Intel already supplied the processors for most PCs, having spent the previous decade successfully fending off AMD, the only other major company producing chips on the x86 instruction set architecture—a foundational set of rules that govern how chips compute—that was the industry standard for PCs. Apple was the only major computer-maker that didn’t use x86-based chips. Now, Jobs and Otellini announced, this would change. Mac computers would have Intel chips inside. Intel’s empire would grow, and its stranglehold on the PC industry would tighten.

乔布斯已经是硅谷的偶像,他发明了麦金塔电脑,并开创了计算机直观易用的理念。2001 年,Apple 发布了 iPod,这是一款富有远见的产品,展示了数字技术如何改变任何消费设备。英特尔的欧德宁与乔布斯截然不同。他被聘为经理,而不是有远见的人。与英特尔之前的首席执行官鲍勃·诺伊斯、戈登·摩尔、安迪·格鲁夫和克雷格·巴雷特不同,欧德宁的背景不是工程或物理学,而是经济学。他毕业于MBA,而不是博士学位。在他担任 CEO 期间,影响力从化学家和物理学家转向经理和会计师。起初几乎察觉不到这一点,尽管员工注意到高管的衬衫越来越白,而且他们更经常打领带。欧德宁继承了一家利润丰厚的公司。他认为他的首要任务是通过以下方式保持尽可能高的利润率榨取英特尔在 x86 芯片上事实上的垄断地位,他运用教科书式的管理实践来捍卫它。

Jobs was already a Silicon Valley icon, having invented the Macintosh and pioneered the idea that computers could be intuitive and easy to use. In 2001, Apple released the iPod, a visionary product showing how digital technology could transform any consumer device. Intel’s Otellini couldn’t have been more different from Jobs. He was hired to be a manager, not a visionary. Unlike Intel’s prior CEOs—Bob Noyce, Gordon Moore, Andy Grove, and Craig Barrett—Otellini’s background was not in engineering or physics, but in economics. He’d graduated with an MBA, not a PhD. His time as CEO saw influence shift from chemists and physicists toward managers and accountants. This was barely perceptible at first, though employees noted that executives’ shirts became steadily whiter and they wore ties more often. Otellini inherited a company that was enormously profitable. He saw his primary task as keeping profit margins as high as possible by milking Intel’s de facto monopoly on x86 chips, and he applied textbook management practices to defend it.

x86 架构主宰 PC 并不是因为它是最好的,而是因为 IBM 的第一台个人计算机碰巧使用了它。与为个人电脑提供操作系统的微软一样,英特尔控制着个人电脑生态系统的这一关键组成部分。这部分是由于运气——IBM 可以为其第一批 PC 选择摩托罗拉的处理器——但也部分归功于安迪·格罗夫的战略远见。在 1990 年代初期的员工会议上,格鲁夫会勾勒出一幅描绘他对计算未来的愿景的图像:一座被护城河环绕的城堡。城堡是英特尔的盈利能力;保卫城堡的护城河是 x86。

The x86 architecture dominated PCs not because it was the best, but because IBM’s first personal computer happened to use it. Like Microsoft, which provided the operating system for PCs, Intel controlled this crucial building block for the PC ecosystem. This was partially by luck—IBM could have chosen Motorola’s processors for its first PCs—but also partly due to Andy Grove’s strategic foresight. At staff meetings in the early 1990s, Grove would sketch an image illustrating his vision of the future of computing: a castle surrounded by a moat. The castle was Intel’s profitability; the moat, defending the castle, was x86.

自英特尔首次采用 x86 架构以来的几年里,伯克利的计算机科学家设计了一种更新、更简单的芯片架构,称为 RISC,它提供了更高效的计算,从而降低了功耗。相比之下,x86 架构复杂且庞大。在 1990 年代,Andy Grove 曾认真考虑将英特尔的主要芯片转换为 RISC 架构,但最终决定放弃。RISC效率更高,但变革成本高,对英特尔事实上的垄断威胁太严重。电脑行业是围绕 x86 设计的,英特尔主导了生态系统。所以 x86 定义了迄今为止的大多数 PC 架构。

In the years since Intel first adopted the x86 architecture, computer scientists at Berkeley had devised a newer, simpler chip architecture called RISC that offered more efficient calculations and thus lower power consumption. The x86 architecture was complex and bulky by comparison. In the 1990s, Andy Grove had seriously considered switching Intel’s main chips to a RISC architecture, but ultimately decided against it. RISC was more efficient, but the cost of change was high, and the threat to Intel’s de facto monopoly was too serious. The computer industry was designed around x86 and Intel dominated the ecosystem. So x86 defines most PC architectures to this day.

英特尔的 x86 指令集架构也主导着服务器业务,随着公司在 2000 年代建立更大的数据中心,以及随后亚马逊网络服务、微软 Azure 和谷歌云等企业构建了创建“云、 ”个人和公司在上面存储数据和运行程序。在 1990 年代和 2000 年代初期,英特尔在为服务器提供芯片的业务中只占有很小的份额,落后于 IBM 和 HP 等公司。但英特尔利用其设计和制造尖端处理器芯片的能力赢得了数据中心市场份额,并将 x86 确立为那里的行业标准。到 2000 年代中期,就在云计算兴起之际,英特尔已经几乎垄断了数据中心芯片,仅与 AMD 竞争。如今,几乎每个主要数据中心都使用英特尔或 AMD 的 x86 芯片。没有处理器,云就无法运行。

Intel’s x86 instruction set architecture also dominates the server business, which boomed as companies built ever larger data centers in the 2000s and then as businesses like Amazon Web Services, Microsoft Azure, and Google Cloud constructed the vast warehouses of servers that create “the cloud,” on which individuals and companies store data and run programs. In the 1990s and early 2000s, Intel had only a small share of the business of providing chips for servers, behind companies like IBM and HP. But Intel used its ability to design and manufacture cutting-edge processor chips to win data center market share and establish x86 as the industry standard there, too. By the mid-2000s, just as cloud computing was emerging, Intel had won a near monopoly over data center chips, competing only with AMD. Today, nearly every major data center uses x86 chips from either Intel or AMD. The cloud can’t function without their processors.

一些公司试图挑战 x86 作为 PC 行业标准的地位。1990 年,Apple 和两个合作伙伴在英国剑桥成立了一家名为 Arm 的合资企业。其目的是使用新的指令集架构设计处理器芯片,该架构基于英特尔曾考虑但拒绝的更简单的 RISC 原则。作为一家初创公司,Arm 没有面临从 x86 转移的成本,因为它没有业务也没有客户。相反,它想取代处于计算生态系统中心的 x86。Arm 的第一任首席执行官 Robin Saxby 对这家 12 人的创业公司抱有远大的抱负。“我们必须成为全球标准,”他告诉他的同事。“那是我们唯一的机会。”

Some companies tried challenging x86’s position as the industry standard in PCs. In 1990, Apple and two partners established a joint venture called Arm, based in Cambridge, England. The aim was to design processor chips using a new instruction set architecture based on the simpler RISC principles that Intel had considered but rejected. As a startup, Arm faced no costs of shifting away from x86, because it had no business and no customers. Instead, it wanted to replace x86 at the center of the computing ecosystem. Arm’s first CEO, Robin Saxby, had vast ambitions for the twelve-person startup. “We have got to be the global standard,” he told his colleagues. “That’s the only chance we’ve got.”

Saxby 曾在摩托罗拉的欧洲半导体部门晋升,之后在一家欧洲芯片初创公司工作,该公司因制造工艺表现不佳而失败。他了解依赖内部制造的局限性。“硅就像钢铁,”他在早期关于 Arm 战略的辩论中坚持说。“这是一种商品…… 我们应该在我的尸体上制造芯片。” 取而代之的是,Arm 采用了一种商业模式,即出售使用其架构的许可证并让任何其他芯片设计师购买它们。这提出了一个分解芯片行业的新愿景。英特尔拥有自己的架构 (x86),在该架构上设计和生产了许多不同的芯片。Saxby 想将他的 Arm 架构出售给无晶圆设计公司,这些公司会根据自己的目的定制 Arm 的架构,然后将制造外包给台积电这样的代工厂。

Saxby had climbed the ranks at Motorola’s European semiconductor divisions before working at a European chip startup that failed because its manufacturing processes underperformed. He understood the limits of relying on in-house manufacturing. “Silicon is like steel,” he insisted in the early debates over Arm’s strategy. “It’s a commodity…. We should build chips over my dead body.” Instead, Arm adopted a business model of selling licenses for use of its architecture and letting any other chip designer buy them. This presented a new vision of a disaggregated chip industry. Intel had its own architecture (x86) on which it designed and produced many different chips. Saxby wanted to sell his Arm architecture to fabless design firms that would customize Arm’s architecture for their own purposes, then outsource the manufacturing to a foundry like TSMC.

Saxby 不仅梦想与英特尔竞争,还梦想颠覆其商业模式。然而,Arm 未能在 1990 年代和 2000 年代赢得 PC 市场份额,因为英特尔与微软的 Windows 操作系统的合作太强大了,无法挑战。然而,Arm 的简化、节能架构很快在必须节省电池使用的小型便携式设备中流行起来。任天堂为其手持视频游戏选择了基于 Arm 的芯片,例如,英特尔从未关注过的一个小市场。英特尔的计算机处理器寡头垄断利润太高,不足以证明考虑利基市场的合理性。英特尔直到为时已晚才意识到它应该在另一个看似利基的便携式计算设备市场竞争:手机。

Saxby didn’t simply dream of rivaling Intel, but of disrupting its business model. However, Arm failed to win market share in PCs in the 1990s and 2000s, because Intel’s partnership with Microsoft’s Windows operating system was simply too strong to challenge. However, Arm’s simplified, energy-efficient architecture quickly became popular in small, portable devices that had to economize on battery use. Nintendo chose Arm-based chips for its handheld video games, for example, a small market that Intel never paid much attention to. Intel’s computer processor oligopoly was too profitable to justify thinking about niche markets. Intel didn’t realize until too late that it ought to compete in another seemingly niche market for a portable computing device: the mobile phone.

移动设备将改变计算的想法并不新鲜。富有远见的加州理工学院教授卡弗·米德(Carver Mead)在 1970 年代初也曾预言过。英特尔也知道,个人电脑不会成为计算发展的最后阶段。该公司在 1990 年代和 2000 年代期间投资了一系列新产品,例如超前二十年的 Zoom-esque 视频会议系统。但这些新产品很少能流行起来,与其说是因为技术原因,不如说是因为它们的利润都远低于英特尔为个人电脑制造芯片的核心业务。他们从未获得英特尔内部的支持。

The idea that mobile devices would transform computing wasn’t new. Carver Mead, the visionary Caltech professor, had predicted as much in the early 1970s. Intel, too, knew that PCs wouldn’t be the final stage in the evolution of computing. The company invested in a series of new products over the course of the 1990s and 2000s, like a Zoom-esque video conferencing system that was two decades ahead of its time. But few of these new products caught on, less for technical reasons than because they were all far less profitable than Intel’s core business of building chips for PCs. They never attracted support from inside Intel.

自 1990 年代初安迪·格鲁夫 (Andy Grove) 仍担任 CEO 以来,移动设备一直是公司经常讨论的话题。1990 年代初期,在英特尔圣克拉拉总部的一次会议上,一位高管在空中挥舞着他的 Palm Pilot 并宣称:“这些设备将成长并取代 PC。” 但是把钱投入的想法移动设备在当时似乎是一场疯狂的赌博卖个人电脑处理器可以赚到钱。所以英特尔决定不进军移动业务,直到为时已晚。

Mobile devices had been a regular source of discussion at the company since the early 1990s, when Andy Grove was still CEO. At one meeting at Intel’s Santa Clara headquarters in the early 1990s, an executive waved his Palm Pilot in the air and declared: “These devices will grow up and replace the PC.” But the idea of pouring money into mobile devices seemed like a wild gamble at a time when there was far more money to be made selling processors for PCs. So Intel decided not to enter the mobile business until it was too late.

为安迪·格鲁夫提供建议的哈佛教授很容易诊断出英特尔的困境。英特尔的每个人都知道 Clayton Christensen 和他的“创新者困境”概念。然而,该公司的 PC 处理器业务看起来很可能在很长一段时间内都印钞票。与 1980 年代,当 Grove 将英特尔从 DRAM 转向,而公司正在亏损时,在 1990 年代和 2000 年代,英特尔是美国最赚钱的公司之一。问题不在于没有人意识到英特尔应该考虑新产品,而在于现状实在太有利可图了。如果英特尔什么都不做,它仍将拥有世界上最有价值的两座城堡——PC 和服务器芯片——被深深的 x86 护城河所包围。

Intel’s dilemma could have been easily diagnosed by the Harvard professor who’d advised Andy Grove. Everyone at Intel knew Clayton Christensen and his concept of “the innovator’s dilemma.” However, the company’s PC processor business looked likely to print money for a very long time. Unlike in the 1980s, when Grove reoriented Intel away from DRAM at a time when the company was bleeding money, in the 1990s and 2000s, Intel was one of America’s most profitable firms. The problem wasn’t that no one realized Intel ought to consider new products, but that the status quo was simply too profitable. If Intel did nothing at all, it would still own two of the world’s most valuable castles—PC and server chips—surrounded by a deep x86 moat.

在将英特尔的芯片用于 Mac 电脑的交易后不久,乔布斯又带着新的推介回到了欧德宁。英特尔会为苹果的最新产品——电脑手机制造芯片吗?所有手机都使用芯片来运行操作系统并管理与手机网络的通信,但苹果希望其手机能够像计算机一样运行。因此,它需要一个强大的计算机式处理器。“他们想付出一定的代价,”欧德宁在事后告诉记者亚历克西斯·马德里加尔,“而不是一分钱……。我看不见。这不是你可以在数量上弥补的这些东西之一。事后看来,预测的成本是错误的音量是任何人想象的 100 倍。” 英特尔拒绝了 iPhone 合同。

Shortly after the deal to put Intel’s chips in Mac computers, Jobs came back to Otellini with a new pitch. Would Intel build a chip for Apple’s newest product, a computerized phone? All cell phones used chips to run their operating systems and manage communication with cell phone networks, but Apple wanted its phone to function like a computer. It would need a powerful computer-style processor as a result. “They wanted to pay a certain price,” Otellini told journalist Alexis Madrigal after the fact, “and not a nickel more…. I couldn’t see it. It wasn’t one of these things you can make up on volume. And in hindsight, the forecasted cost was wrong and the volume was 100× what anyone thought.” Intel turned down the iPhone contract.

苹果在别处寻找手机芯片。乔布斯转向 Arm 的架构,与 x86 不同,它针对必须节省功耗的移动设备进行了优化。早期的 iPhone 处理器是由三星生产的,三星是追随台积电进入代工业务的。欧德宁关于 iPhone 将成为小众产品的预测被证明是非常错误的。然而,当他意识到自己的错误时,为时已晚。英特尔后来争相赢得智能手机业务的份额。尽管最终在智能手机产品上投入了数十亿美元,但英特尔从未有过太多的表现。在欧德宁和英特尔意识到发生了什么之前,苹果在其利润丰厚的城堡周围挖了一条深沟。

Apple looked elsewhere for its phone chips. Jobs turned to Arm’s architecture, which unlike x86 was optimized for mobile devices that had to economize on power consumption. The early iPhone processors were produced by Samsung, which had followed TSMC into the foundry business. Otellini’s prediction that the iPhone would be a niche product proved horribly wrong. By the time he realized his mistake, however, it was too late. Intel would later scramble to win a share of the smartphone business. Despite eventually pouring billions of dollars into products for smartphones, Intel never had much to show for it. Apple dug a deep moat around its immensely profitable castle before Otellini and Intel realized what was happening.

就在英特尔拒绝 iPhone 合同的几年后,苹果在智能手机上赚的钱比英特尔卖的 PC 处理器还多。英特尔多次尝试翻越苹果城堡的围墙,但已经失去了先发优势。花费数十亿获得第二名几乎没有吸引力,尤其是在英特尔的个人电脑业务仍然高利润并且其数据中心业务增长迅速的情况下。所以英特尔从未找到在移动设备中赢得立足点的方法,如今移动设备消耗了近三分之一的已售芯片。它仍然没有。

Just a handful of years after Intel turned down the iPhone contract, Apple was making more money in smartphones than Intel was selling PC processors. Intel tried several times to scale the walls of Apple’s castle but had already lost first-mover advantage. Spending billions for second place was hardly appealing, especially since Intel’s PC business was still highly profitable and its data center business was growing quickly. So Intel never found a way to win a foothold in mobile devices, which today consume nearly a third of chips sold. It still hasn’t.

格鲁夫离开后的这些年里,英特尔错失良机都有一个共同的原因。自 1980 年代后期以来,英特尔已经获得了 25 万亿美元的利润,甚至在调整通货膨胀之前,这一记录很少有其他公司能够与之匹敌。它通过为 PC 和服务器芯片收取大量费用来做到这一点。由于 Grove 磨练并留给继任者的优化设计流程和先进制造技术,英特尔能够维持高价格。该公司的领导层始终优先考虑生产利润率最高的芯片。

Intel’s missed opportunities in the years since Grove left the scene all had a common cause. Since the late 1980s, Intel has made a quarter trillion dollars in profit, even before adjusting for inflation, a track record that few other companies have matched. It has done this by charging a ton for PC and server chips. Intel could sustain high prices because of the optimized design processes and advanced manufacturing that Grove had honed and bequeathed to his successors. The company’s leadership consistently prioritized the production of chips with the highest profit margin.

这是一种理性的策略——没有人想要利润率低的产品——但这使得尝试任何新事物变得不可能。对实现短期利润目标的执着开始取代长期的技术领先地位。从工程师到管理人员的权力转移加速了这一过程。2005 年至 2013 年担任英特尔首席执行官的欧德宁承认,由于担心财务问题,他拒绝了制造 iPhone 芯片的合同。对利润率的执着深深地渗入了公司——它的招聘决策、产品路线图和研发流程。该公司的领导者只是更专注于设计公司的资产负债表,而不是晶体管。“它有技术,有人才,”英特尔的一位前财务主管回忆道。“这只是不想受到利润的冲击。”

This was a rational strategy—no one wants products with low profit margins—but it made it impossible to try anything new. A fixation on hitting short-term margin targets began to replace long-term technology leadership. The shift in power from engineers to managers accelerated this process. Otellini, Intel’s CEO from 2005 to 2013, admitted he turned down the contract to build iPhone chips because he worried about the financial implications. A fixation on profit margins seeped deep into the firm—its hiring decisions, its product road maps, and its R&D processes. The company’s leaders were simply more focused on engineering the company’s balance sheet than its transistors. “It had the technology, it had the people,” one former finance executive at Intel reminisced. “It just didn’t want to take the margin hit.”

第 34 章 跑得更快?

CHAPTER 34 Running Faster?

2010 年,安迪·格罗夫( Ndy Grove) 在帕洛阿尔托 (Palo Alto) 的一家餐厅用餐时,被介绍给了三位正在硅谷巡回演出的中国风险投资家。他于 2005 年辞去英特尔董事长一职,现在是一名简单的退休人员。他建立然后拯救的公司仍然非常有利可图。它甚至在 2008 年和 2009 年都赚钱,尽管硅谷的失业率飙升至 9% 以上。然而,格鲁夫并不认为英特尔过去的成功是自满的理由。他一如既往地偏执。看到中国风险资本家在帕洛阿尔托投资让他想知道:硅谷在大规模失业的情况下将生产外包是否明智?

Andy Grove was dining at a Palo Alto restaurant in 2010 when he was introduced to three Chinese venture capitalists who were touring Silicon Valley. He’d stepped down as Intel’s chairman in 2005 and was now a simple retiree. The company he’d built and then rescued was still immensely profitable. It made money even in 2008 and 2009, though Silicon Valley’s unemployment rate spiked above 9 percent. However, Grove didn’t view Intel’s past success as an argument for complacency. He was as paranoid as ever. Seeing Chinese venture capitalists investing in Palo Alto made him wonder: Was Silicon Valley smart to be offshoring production at a time of mass unemployment?

作为纳粹和苏联军队的犹太难民,格鲁夫不是本土主义者。英特尔从世界各地聘请工程师。它在多个大洲经营设施。然而,格鲁夫担心先进制造业工作的外包。仅在三年前推出的 iPhone 就是这一趋势的例证。很少有 iPhone 的组件是在美国制造的 尽管离岸外包始于低技能工作,但格鲁夫认为它不会止步于此,无论是在半导体还是任何其他行业。他担心电动汽车所需的锂电池,尽管美国发明了大部分锂电池,但在该市场的份额很小。核心技术。他的解决方案是:“对离岸劳动力的产品征收额外税。如果结果是贸易战,那就像对待其他战争一样对待它——为胜利而战。”

As a Jewish refugee from Nazi and Soviet armies, Grove was no nativist. Intel hired engineers from the world over. It operated facilities on multiple continents. However, Grove was worried about the offshoring of advanced manufacturing jobs. The iPhone, which had been introduced just three years earlier, exemplified the trend. Few of the iPhone’s components were built in the U.S. Though offshoring started with low-skilled jobs, Grove didn’t think it would stop there, whether in semiconductors or any other industry. He worried about lithium batteries needed for electric vehicles, where the U.S. made up a tiny share of the market despite having invented much of the core technology. His solution: “Levy an extra tax on the product of offshored labor. If the result is a trade war, treat it like other wars—fight to win.”

许多人选择将格鲁夫视为过去时代的代表。在互联网出现之前,他早在一代人之前就建立了英特尔。他的公司错过了手机,靠x86垄断的果实为生。在 2010 年代初期,英特尔保留了世界上最先进的半导体工艺技术,在竞争对手之前推出了更小的晶体管,其节奏与自戈登·摩尔时代以来众所周知的相同。然而,英特尔与台积电、三星等竞争对手的差距已经开始缩小。

Many people chose to write off Grove as a representative of a bygone era. He’d built Intel a generation earlier, before the internet existed. His company missed the mobile phone and was living off the fruits of its x86 monopoly. In the early 2010s, Intel retained the world’s most advanced semiconductor process technology, introducing smaller transistors before rivals, with the same regular cadence it had been known for since the days of Gordon Moore. However, the gap between Intel and rivals like TSMC and Samsung had begun to shrink.

此外,英特尔的业务现在被其他具有不同商业模式的科技公司所掩盖。英特尔在 2000 年代初曾是世界上最有价值的公司之一,但已被苹果超越,后者的新移动生态系统并不依赖英特尔的芯片。英特尔错过了互联网经济的兴起。Facebook 成立于 2006 年,到 2010 年其市值几乎是英特尔的一半。它很快就会变得有价值几倍。硅谷最大的芯片制造商可能会反驳说,互联网数据是在其服务器芯片上处理的,并在依赖其处理器的个人电脑上访问。然而,生产芯片的利润不如在应用程序上销售广告。格鲁夫崇拜“颠覆性创新”,但到 2010 年代,英特尔的业务被打乱了。他对苹果离岸装配线的哀叹被置若罔闻。

Moreover, Intel’s business was now overshadowed by other tech firms with different business models. Intel had been one of the world’s most valuable companies in the early 2000s, but had been overtaken by Apple, whose new mobile ecosystem didn’t rely on Intel’s chips. Intel missed the rise of the internet economy. Facebook, founded in 2006, was by 2010 worth nearly half as much as Intel. It would soon become several times more valuable. The Valley’s biggest chipmaker could retort that the internet’s data was processed on its server chips and accessed on PCs reliant on its processors. Yet producing chips was less profitable than selling ads on apps. Grove idolized “disruptive innovation,” but by the 2010s, Intel’s business was being disrupted. His lament of Apple’s offshored assembly lines fell on deaf ears.

即使在半导体领域,格鲁夫充满厄运的预言也被广泛拒绝。诚然,像台积电这样的新半导体代工厂大部分都在海外。然而,外国代工厂生产的芯片主要由美国无晶圆厂公司设计。此外,他们的工厂里到处都是美国制造的制造设备。自从仙童半导体(Andy Grove 的第一个雇主)在香港开设了最初的组装厂以来,向东南亚外包一直是芯片行业商业模式的核心。

Even in the semiconductor space, Grove’s doom-filled prophesies were widely rejected. True, new semiconductor foundries like TSMC were largely offshore. Yet foreign foundries produced chips largely designed by American fabless firms. Moreover, their fabs were full of U.S.-made manufacturing equipment. Offshoring to Southeast Asia had been central to the chip industry’s business model since Fairchild Semiconductor—Andy Grove’s first employer—opened its initial assembly plant in Hong Kong.

格鲁夫不相信。“放弃今天的‘大宗商品’制造业可以将你排除在明天的新兴产业之外,”他宣称,指向电池行业。格罗夫写道,美国“在 30 年前停止生产消费电子设备时,就失去了在电池方面的领先地位”。然后它错过了PC电池,现在远远落后于电动汽车电池。“我怀疑他们会赶上,”他在 2010 年预测。

Grove wasn’t convinced. “Abandoning today’s ‘commodity’ manufacturing can lock you out of tomorrow’s emerging industry,” he declared, pointing to the electric battery industry. The U.S. “lost its lead in batteries thirty years ago when it stopped making consumer electronics devices,” Grove wrote. Then it missed PC batteries, and now was far behind on batteries for electric vehicles. “I doubt they will ever catch up,” he predicted in 2010.

即使在半导体行业,也很容易找到与 Grove 对离岸专业知识的悲观态度的对应物。与 1980 年代后期日本竞争对手在 DRAM 设计和制造方面击败硅谷的情况相比,美国的芯片生态系统看起来更健康。印刷巨额利润的不仅是英特尔。许多无晶圆厂芯片设计师也是如此。除了尖端光刻技术的流失外,美国的半导体制造设备公司在 2000 年代普遍蓬勃发展。应用材料公司仍然是世界上最大的半导体工具制造公司,制造设备,例如在加工硅晶片时将化学薄膜沉积在硅晶片上的机器。Lam Research 在将电路蚀刻到硅晶圆上方面拥有世界一流的专业知识。和 KLA,也位于硅谷,拥有世界上最好的工具来发现晶圆和光刻掩模上的纳米级错误。这三个工具制造商正在推出可以在原子尺度上沉积、蚀刻和测量特征的新一代设备,这对于制造下一代芯片至关重要。几家日本公司——尤其是东京电子公司——具有与美国设备制造商相当的能力。然而,如果不使用一些美国工具,基本上是不可能做出领先的芯片的。几家日本公司——尤其是东京电子公司——具有与美国设备制造商相当的能力。然而,如果不使用一些美国工具,基本上是不可能做出领先的芯片的。几家日本公司——尤其是东京电子公司——具有与美国设备制造商相当的能力。然而,如果不使用一些美国工具,基本上是不可能做出领先的芯片的。

Even within the semiconductor industry, it was easy to find counterpoints to Grove’s pessimism about offshoring expertise. Compared to the situation in the late 1980s, when Japanese competitors were beating Silicon Valley in terms of DRAM design and manufacturing, America’s chip ecosystem looked healthier. It wasn’t only Intel that was printing immense profits. Many fabless chip designers were, too. Except for the loss of cutting-edge lithography, America’s semiconductor manufacturing equipment firms generally thrived during the 2000s. Applied Materials remained the world’s largest semiconductor toolmaking company, building equipment like the machines that deposited thin films of chemicals on top of silicon wafers as they were processed. Lam Research had world-beating expertise in etching circuits into silicon wafers. And KLA, also based in Silicon Valley, had the world’s best tools for finding nanometer-sized errors on wafers and lithography masks. These three toolmakers were rolling out new generations of equipment that could deposit, etch, and measure features at the atomic scale, which would be crucial for making the next generation of chips. A couple Japanese firms—notably, Tokyo Electron—had some comparable capabilities to America’s equipment makers. Nevertheless, it was basically impossible to make a leading-edge chip without using some American tools.

设计芯片也是如此。到 2010 年代初,最先进的微处理器已经每个芯片上有十亿个晶体管。能够布置这些晶体管的软件由三个美国公司提供,Cadence、Synopsys 和 Mentor,它们控制了约四分之三的市场。如果不使用这些公司的至少一个软件,就不可能设计芯片。此外,大多数提供芯片设计软件的小公司也位于美国。没有其他国家能接近。

The same was true for designing chips. By the early 2010s, the most advanced microprocessors had a billion transistors on each chip. The software capable of laying out these transistors was provided by three American firms, Cadence, Synopsys, and Mentor, which controlled around three-quarters of the market. It was impossible to design a chip without using at least one of these firms’ software. Moreover, most of the smaller firms providing chip design software were U.S.-based, too. No other country came close.

当华尔街和华盛顿的分析师审视硅谷时,他们看到了一个有利可图且技术进步的芯片行业。当然,如此严重地依赖台湾的几家工厂来生产世界上很大一部分芯片是有一些风险的。1999 年,一场里氏 7.3 级的地震袭击了台湾,导致该国大部分地区断电,包括两座核电站。台积电的晶圆厂也断电,威胁到该公司的生产和全球许多芯片。

When analysts on Wall Street and in Washington looked at Silicon Valley, they saw a chip industry that was profitable and advancing technologically. There were, of course, some risks of relying so heavily on a couple of facilities in Taiwan to manufacture a large share of the world’s chips. In 1999, an earthquake measuring 7.3 on the Richter scale struck Taiwan, knocking out power across much of the country, including from two nuclear power plants. TSMC’s fabs lost power, too, threatening the company’s production and many of the world’s chips.

张忠谋迅速与台湾官员通电话,以确保该公司获得优先用电。公司五家晶圆厂中的四家花了一周时间才重新上线。第五次花了更长的时间。然而,中断是有限的,消费电子市场在一个月内恢复正常。然而,1999 年的地震只是该岛在 20 世纪遭受的第三强地震;很容易想象更强烈的地震冲击。台积电的客户被告知,该公司的设施可以承受里氏 9 级的地震,其中世界经历过自 1900 年以来的五个。这不是任何人都想测试的说法。然而,台积电总能指出,硅谷位于圣安地列斯断层之上,因此将制造业带回加州并不安全。

Morris Chang was quickly on the phone with Taiwanese officials to ensure the company got preferential access to electricity. It took a week to get four of the company’s five fabs back online; the fifth took even longer. However, disruptions were limited and the market for consumer electronics reverted to normal within a month. However, the 1999 earthquake was only the third strongest the island had suffered in the twentieth century; it was easy to imagine stronger seismic shocks. TSMC’s customers were told that the company’s facilities could tolerate earthquakes measuring 9 on the Richter scale, of which the world has experienced five since 1900. This was not a claim that anyone wanted to test. However, TSMC could always point out that Silicon Valley sat atop the San Andreas Fault, so bringing manufacturing back to California wasn’t much safer.

一个更困难的问题是,美国政府应如何调整其对半导体技术外国销售的控制,以适应日益国际化的供应链。除了为美国军方生产专用半导体的几家小型芯片制造商外,硅谷巨头在 1990 年代和 2000 年代降低了与五角大楼的关系。当他们在 1980 年代面对日本的竞争时,硅谷的 CEO 们在国会大厅里度过了很多时间。现在他们认为他们不需要政府的帮助。他们主要担心政府会通过与其他国家签署贸易协议并取消对出口的控制来让路。华盛顿的许多官员支持该行业要求放松控制的呼吁。中国有像中芯国际这样雄心勃勃的公司,但华盛顿的共识是贸易和投资将鼓励中国成为正如有影响力的外交官罗伯特佐利克所说,国际体系的“负责任的利益相关者”。

A more difficult question was how the U.S. government should adjust its controls on foreign sales of semiconductor technology to account for an increasingly international supply chain. Except for a couple of small chipmakers that produced specialized semiconductors for the U.S. military, Silicon Valley giants downgraded their relations with the Pentagon during the 1990s and 2000s. When they’d faced Japanese competition in the 1980s, Silicon Valley CEOs spent plenty of time in the halls of Congress. Now they didn’t think they needed government help. Their main concern was for government to get out of the way, by signing trade deals with other countries and removing controls on exports. Many officials in Washington backed the industry’s calls for looser controls. China had ambitious companies like SMIC, but the consensus in Washington was that trade and investment would encourage China to become a “responsible stakeholder” of the international system, as influential diplomat Robert Zoellick put it.

此外,关于全球化的流行理论使实施严格控制听起来几乎是不可能的。冷战期间,控制措施难以实施,引发了美国和盟国之间关于哪些设备可以卖给苏联的经常性争论。与苏联不同,2000 年代的中国更加融入世界经济。华盛顿得出的结论是,出口管制弊大于利,既损害了美国工业,又不妨碍中国从其他国家的公司购买商品。日本和欧洲急于向中国出售几乎所有东西。华盛顿没有人愿意与盟友就出口管制进行斗争,尤其是在美国领导人专注于与中国同行交朋友的情况下。

Moreover, popular theories about globalization made it sound almost impossible to impose strict controls. Controls had been hard enough to enforce during the Cold War, sparking regular disputes between the U.S. and allies about what equipment could be sold to the Soviets. Unlike the USSR, China in the 2000s was far more integrated into the world economy. Washington concluded that export controls would do more harm than good, hurting U.S. industry without preventing China from buying goods from firms in other countries. Japan and Europe were eager to sell almost anything to the PRC. No one in Washington had the stomach for a fight with allies about export controls, especially as U.S. leaders were focused on befriending their Chinese counterparts.

华盛顿围绕这样一种观点形成了新的共识,即最好的政策是比美国的竞争对手“跑得更快”。“对于任何一种产品,尤其是半导体,美国将越来越依赖任何一个国家(更不用说中国)的可能性是非常小,”一位美国专家预测道。美国竟然给予中国中芯国际特殊地位“经过验证的最终用户”,证明该公司没有向中国军方出售产品,因此免于某些出口管制。除了少数立法者——大多数是南方共和党人,他们仍然把中国视为冷战从未结束——华盛顿几乎每个人都支持这一战略。比竞争对手“跑得更快”。

A new consensus in Washington formed around the idea that the best policy was to “run faster” than America’s rivals. “The likelihood that the United States will grow dependent on any one country, much less China, for any one product, especially semiconductors, is exceedingly small,” predicted one American expert. The U.S. went so far as to give China’s SMIC special status as a “validated end-user,” certifying that the company didn’t sell to the Chinese military and was thus exempt from certain export controls. Other than a handful of legislators—mostly Southern Republicans who still looked at China as though the Cold War had never ended—almost everyone in Washington backed the strategy of “running faster” than rivals.

“跑得更快”是一个优雅的策略,只有一个问题:从一些关键指标来看,美国并没有跑得更快,它正在失去阵地。政府中几乎没有人费心去做分析,但安迪·格鲁夫关于专业知识离岸外包的悲观预测部分实现了。2007 年,国防部委托五角大楼前官员理查德·范·阿塔和几位同事进行了一项研究,以评估半导体行业“全球化”对军方供应链的影响。范阿塔从事防守工作几十年来,微电子经历了日本芯片产业的兴衰。他不容易反应过度,并且了解跨国供应链如何提高行业效率。在和平时期,这个系统运行良好。然而,五角大楼不得不考虑最坏的情况。Van Atta 报告说,国防部获得尖端芯片的机会很快将取决于外国,因为如此多的先进制造正在转移到国外。

“Run faster” was an elegant strategy with only a single problem: by some key metrics, the U.S. wasn’t running faster, it was losing ground. Hardly anyone in government bothered to do the analysis, but Andy Grove’s gloomy predictions about the offshoring of expertise were partially coming true. In 2007, the Defense Department commissioned a study from former Pentagon official Richard Van Atta and several colleagues to assess the impact of semiconductor industry “globalization” on the military’s supply chains. Van Atta had worked on defense microelectronics for several decades and had lived through the rise and fall of Japan’s chip industry. He wasn’t prone to overreaction and understood how a multinational supply chain made the industry more efficient. In peacetime, this system worked smoothly. However, the Pentagon had to think about worst-case scenarios. Van Atta reported that the Defense Department’s access to cutting-edge chips would soon depend on foreign countries because so much advanced fabrication was moving abroad.

在美国单极时刻的狂妄自大中,几乎没有人愿意倾听。华盛顿的大多数人只是简单地得出结论,美国“跑得更快”,甚至没有看一眼证据。然而,半导体行业的历史并没有表明美国的领导地位得到保证。美国在 1980 年代并没有超过日本,尽管它在 1990 年代做到了。GCA 在光刻方面并没有超过尼康或 ASML。美光是唯一能够与东亚竞争对手并驾齐驱的 DRAM 生产商,而许多其他美国 DRAM 生产商却破产了。到 2000 年代末,英特尔在生产微型晶体管方面保持领先于三星和台积电,但差距已经缩小。英特尔的运行速度较慢,尽管它仍然受益于其更先进的起点。美国是大多数类型芯片设计的领导者,尽管台湾的联发科正在证明其他国家也可以设计芯片。范·阿塔几乎没有看到自信的理由,也没有看到自满的理由。“美国的领导地位,”他在 2007 年警告说,“未来十年可能会严重削弱。”没有人在听。

Amid the hubris of America’s unipolar moment, hardly anyone was willing to listen. Most people in Washington simply concluded the U.S. was “running faster” without even glancing at the evidence. However, the history of the semiconductor industry didn’t suggest that U.S. leadership was guaranteed. America hadn’t outrun the Japanese in the 1980s, though it did in the 1990s. GCA hadn’t outrun Nikon or ASML in lithography. Micron was the only DRAM producer able to keep pace with East Asian rivals, while many other U.S. DRAM producers went bust. Through the end of the 2000s, Intel retained a lead over Samsung and TSMC in producing miniaturized transistors, but the gap had narrowed. Intel was running more slowly, though it still benefitted from its more advanced starting point. The U.S. was a leader in most types of chip design, though Taiwan’s MediaTek was proving that other countries could design chips, too. Van Atta saw few reasons for confidence and none for complacency. “The U.S. leadership position,” he warned in 2007, “will likely erode seriously over the next decade.” No one was listening.

第六部分 离岸创新?

PART VI OFFSHORING INNOVATION?

第 35 章 “真正的男人有工厂”

CHAPTER 35 “Real Men Have Fabs”

杰瑞·桑德斯( J erry Sanders)是劳力士(Rolex)的创始人、劳斯莱斯(Rolls Royce)的创始人,他喜欢将拥有一家半导体工厂与在您的游泳池里放一条宠物鲨鱼进行比较。鲨鱼的饲养成本很高,需要时间和精力来维持,并且最终可能会杀了你。尽管如此,桑德斯仍然确信一件事:他永远不会放弃他的晶圆厂。尽管他在伊利诺伊大学读本科时学习过电气工程,但他从来都不是制造业的人。他在飞兆半导体的销售和市场营销方面不断提升,使他成为公司最优秀的人华丽而成功的推销员。

Jerry Sanders, the Rolex-clad, Rolls Royce−driving brawler who founded AMD, liked to compare owning a semiconductor fab with putting a pet shark in your swimming pool. Sharks cost a lot to feed, took time and energy to maintain, and could end up killing you. Even still, Sanders was sure of one thing: he’d never give up his fabs. Though he had studied electrical engineering as an undergraduate at the University of Illinois, he was never a manufacturing guy. He moved up the ranks in sales and marketing at Fairchild Semiconductor, making his name as the company’s most flamboyant and successful salesman.

他的专长是销售,但桑德斯从未梦想过放弃 AMD 的制造设施,尽管台积电等代工厂的崛起使得大型芯片公司有可能考虑将其制造业务和外包给亚洲的代工厂。在 1980 年代与日本争夺 DRAM 市场份额并在 1990 年代与英特尔争夺 PC 市场后,Sanders 致力于他的晶圆厂。他认为它们对 AMD 的成功至关重要。

His specialty was sales, but Sanders never dreamed of giving up AMD’s manufacturing facilities, even as the rise of foundries like TSMC made it possible for big chip firms to consider divesting their manufacturing operations and outsourcing to a foundry in Asia. Having brawled with the Japanese for DRAM market share in the 1980s and with Intel for the PC market in the 1990s, Sanders was committed to his fabs. He thought they were crucial to AMD’s success.

不过,即使是他也承认,在拥有和经营晶圆厂的同时赚钱变得越来越难。问题很简单:每一代的技术改进都让晶圆厂变得更加昂贵。早在几十年前,张忠谋就得出了类似的结论,这也是他认为台积电的商业模式优越的原因。像台积电这样的代工厂可以为许多芯片设计师制造芯片,从其大量生产中榨取效率,而其他公司很难复制。

Even he admitted, though, that it was becoming harder to make money while owning and operating a fab. The problem was simple: each generation of technological improvement made fabs more expensive. Morris Chang had drawn a similar conclusion several decades earlier, which is why he thought TSMC’s business model was superior. A foundry like TSMC could fabricate chips for many chip designers, wringing out efficiencies from its massive production volumes that other companies would find difficult to replicate.

并非芯片行业的所有部门都面临类似的动态,但许多部门确实如此。到 2000 年代,半导体行业通常分为三类。“逻辑”是指运行智能手机、计算机和服务器的处理器。“内存”指的是 DRAM,它提供计算机运行所需的短期内存,以及闪存,也称为 NAND,它随着时间的推移记住数据。第三类芯片更为分散,包括模拟芯片,如将视觉或音频信号转换为数字数据的传感器、与手机网络通信的射频芯片,以及管理设备如何使用电力的半导体。

Not all sectors of the chip industry faced similar dynamics, but many did. By the 2000s, it was common to split the semiconductor industry into three categories. “Logic” refers to the processors that run smartphones, computers, and servers. “Memory” refers to DRAM, which provides the short-term memory computers need to operate, and flash, also called NAND, which remembers data over time. The third category of chips is more diffuse, including analog chips like sensors that convert visual or audio signals into digital data, radio frequency chips that communicate with cell phone networks, and semiconductors that manage how devices use electricity.

这第三类主要不是依靠摩尔定律来推动性能改进。巧妙的设计比缩小晶体管更重要。今天,大约四分之三的此类芯片是在处理器上生产的,处理器速度或大于180 纳米,一种在 1990 年代后期首创的制造技术。因此,这一领域的经济性不同于逻辑和存储芯片,后者必须不断缩小晶体管才能保持领先地位。这类芯片的晶圆厂通常不需要每隔几年就竞相使用最小的晶体管,因此平均而言它们要便宜得多需要四分之一的资本投资用于逻辑或存储芯片的先进晶圆厂。今天,最大的模拟芯片制造商是美国、欧洲或日本。他们的大部分生产也发生在这三个地区,只有一小部分离岸到台湾和韩国。今天最大的模拟芯片制造商是德州仪器,它未能在个人电脑、数据中心或智能手机生态系统中建立英特尔式的垄断地位,但仍然是一家拥有大量模拟芯片和传感器的中等规模、高利润的芯片制造商。现在还有许多其他美国模拟芯片制造商,如 Onsemi、Skyworks 和 Analog Devices,以及欧洲和日本的同类公司。

This third category has not been primarily dependent on Moore’s Law to drive performance improvements. Clever design matters more than shrinking transistors. Today around three-quarters of this category of chips are produced on processors at or larger than 180 nanometers, a manufacturing technology that was pioneered in the late 1990s. As a result, the economics of this segment are different from logic and memory chips that must relentlessly shrink transistors to remain on the cutting edge. Fabs for these types of chips generally don’t need to race toward the smallest transistors every couple of years, so they’re substantially cheaper, on average requiring a quarter the capital investment of an advanced fab for logic or memory chips. Today, the biggest analog chipmakers are American, European, or Japanese. Most of their production occurs in these three regions, too, with only a sliver offshored to Taiwan and South Korea. The largest analog chipmaker today is Texas Instruments, which failed to establish an Intel-style monopoly in the PC, data center, or smartphone ecosystems but remains a medium-sized, highly profitable chipmaker with a vast catalog of analog chips and sensors. There are many other U.S.-based analog chipmakers now, like Onsemi, Skyworks, and Analog Devices, alongside comparable companies in Europe and Japan.

相比之下,内存市场一直被无情地推动将生产外包给少数几家工厂,主要是在东亚。两种主要类型的内存芯片——DRAM 和 NAND——不是以发达经济体为中心的分散供应商,而是仅由几家公司生产。对于 DRAM 存储芯片(1980 年代硅谷与日本发生冲突的半导体类型)而言,一座先进的晶圆厂可能耗资 200 亿美元。曾经有几十家DRAM生产商,但今天只有三个主要生产商。在 1990 年代后期,日本几家陷入困境的 DRAM 生产商合并为一家名为 Elpida 的公司,该公司试图与爱达荷州的美光以及韩国的三星和 SK 海力士竞争。到 2000 年代末,这四家公司控制了大约 85% 的市场。然而尔必达挣扎求生,2013 年被美光收购。与三星和海力士在韩国生产大部分 DRAM 不同,美光的一连串收购使其在日本、台湾、新加坡以及美国拥有 DRAM 晶圆厂。新加坡等国家的政府补贴鼓励美光在当地维持和扩大晶圆厂产能。因此,即使一家美国公司是全球三大 DRAM 生产商之一,但大多数 DRAM 制造商都在东亚。

The memory market, by contrast, has been dominated by a relentless push toward offshoring production to a handful of facilities, mostly in East Asia. Rather than a diffuse set of suppliers centered in advanced economies, the two main types of memory chip—DRAM and NAND—are produced by only a couple of firms. For DRAM memory chips, the type of semiconductor that defined Silicon Valley’s clash with Japan in the 1980s, an advanced fab can cost $20 billion. There used to be dozens of DRAM producers, but today there are only three major producers. In the late 1990s, several of Japan’s struggling DRAM producers were consolidated into a single company, called Elpida, which sought to compete with Idaho’s Micron and with Korea’s Samsung and SK Hynix. By the end of the 2000s, these four companies controlled around 85 percent of the market. Yet Elpida struggled to survive and in 2013 was bought by Micron. Unlike Samsung and Hynix, which produce most of their DRAM in South Korea, Micron’s long string of acquisitions left it with DRAM fabs in Japan, Taiwan, and Singapore as well as in the United States. Government subsidies in countries like Singapore encouraged Micron to maintain and expand fab capacity there. So even though an American company is one of the world’s three biggest DRAM producers, most DRAM manufacturing is in East Asia.

另一种主要存储芯片类型 NAND 的市场也以亚洲为中心。三星,最大的玩家,供应 35% 的市场,其余由韩国海力士、日本铠侠和两家美国公司——美光和西部数据生产。韩国公司几乎完全在韩国或中国生产芯片,但美光和西部数据的 NAND 生产只有一部分在美国,其他生产在新加坡和日本。与 DRAM 一样,虽然美国公司在 NAND 生产中发挥着重要作用,但美国制造的份额要低得多。

The market for NAND, the other main type of memory chip, is also Asia-centric. Samsung, the biggest player, supplies 35 percent of the market, with the rest produced by Korea’s Hynix, Japan’s Kioxia, and two American firms—Micron and Western Digital. The Korean firms produce chips almost exclusively in Korea or China, but only a portion of Micron and Western Digital’s NAND production is in the U.S., with other production in Singapore and Japan. As with DRAM, while U.S. firms play a major role in NAND production, the share of U.S.-based fabrication is substantially lower.

然而,美国在存储芯片产量方面的二流地位并不是什么新鲜事。它可以追溯到 1980 年代后期,当时日本的 DRAM 产量首次超过美国。近年来的重大转变是美国生产的逻辑芯片份额的崩溃。今天,建一个先进的逻辑工厂耗资 200 亿美元,这是一项很少有公司能够承受的巨额资本投资。与存储芯片一样,公司生产的芯片数量与其产量(实际工作的芯片数量)之间存在相关性。鉴于规模优势,制造先进逻辑芯片的公司数量不断减少。

America’s second-rate status in memory chip output, however, is nothing new. It dates to the late 1980s, when Japan first overtook the U.S. in DRAM output. The big shift in recent years is the collapse in the share of logic chips produced in the United States. Today, building an advanced logic fab costs $20 billion, an enormous capital investment that few firms can afford. As with memory chips, there’s a correlation between the number of chips a firm produces and its yield—the number of chips that actually work. Given the benefits of scale, the number of firms fabricating advanced logic chips has shrunk relentlessly.

除了英特尔之外,许多美国主要的逻辑芯片制造商已经放弃了他们的晶圆厂并外包制造。其他以前的主要参与者,如摩托罗拉或美国国家半导体,破产、被收购或市场份额缩小。它们被无晶圆厂公司所取代,这些公司通常从传统半导体公司聘请芯片设计师,但将制造外包给台积电或亚洲的其他代工厂。这让无晶圆厂公司专注于他们的优势——芯片设计——而不需要同时具备制造半导体的专业知识。

With the prominent exception of Intel, many key American logic chipmakers have given up their fabs and outsourced manufacturing. Other formerly major players, like Motorola or National Semiconductor, went bankrupt, were purchased, or saw their market share shrink. They were replaced by fabless firms, which often hired chip designers from legacy semiconductor firms but outsourced fabrication to TSMC or other foundries in Asia. This let fabless companies focus on their strength—chip design—without requiring simultaneous expertise in fabricating semiconductors.

只要桑德斯担任首席执行官,他创立的公司 AMD 就一直从事制造逻辑芯片的业务,比如 PC 处理器。老派硅谷的 CEO 们一直坚持认为,将半导体制造与设计分开会导致效率低下。但是,使芯片设计和芯片制造集成了这么长时间的是文化,而不是商业推理。桑德斯还记得鲍勃·诺伊斯在仙童实验室修修补补的日子。他支持将 AMD 的制造保留在内部的论点依赖于很快过时的男子气概的姿态。当他在 1990 年代听到一位记者的俏皮话“真正的男人有工厂”时,他采用了这个短语作为自己的短语。“现在听我说,好好听我说,”桑德斯在一次行业会议上宣称。“真正的男人有工厂。”

So long as Sanders was CEO, AMD, the company he founded, stayed in the business of manufacturing logic chips, like processors for PCs. Old-school Silicon Valley CEOs kept insisting that separating the fabrication of semiconductors from their design caused inefficiencies. But it was culture, not business reasoning, that kept chip design and chip fabrication integrated for so long. Sanders could still remember the days of Bob Noyce tinkering away in Fairchild’s lab. His argument in favor of keeping AMD’s manufacturing in-house relied on macho-man posturing that was quickly going out of date. When he heard a quip from a journalist in the 1990s that “real men have fabs,” he adopted the phrase as his own. “Now hear me and hear me well,” Sanders declared at one industry conference. “Real men have fabs.”

第 36 章 无晶圆厂革命

CHAPTER 36 The Fabless Revolution

真正的男人”可能拥有晶圆厂,但硅谷的新一波半导体企业家却没有。自 1980 年代后期以来,无晶圆厂芯片公司的数量呈爆炸式增长,这些公司在内部设计半导体但外包制造,通常依赖台积电提供这项服务。当 Gordon Campbell 和 Dado Banatao 于 1984 年创立 Chips and Technologies(通常被认为是第一家无晶圆厂公司)时,一位朋友声称“不是一家真正的半导体公司”,因为它没有制造自己的芯片。然而,事实证明,他们为个人电脑设计的图形芯片很受欢迎,与一些业内最大厂商的产品竞争。最终,芯片和技术公司衰落并被英特尔收购。然而,事实证明,无晶圆厂的商业模式是可行的,只需要一个好主意和几百万美元的启动资金,这只是建造晶圆厂所需资金的一小部分。

“Real men” might have fabs, but Silicon Valley’s new wave of semiconductor entrepreneurs didn’t. Since the late 1980s, there’s been explosive growth in the number of fabless chip firms, which design semiconductors in-house but outsource their manufacturing, commonly relying on TSMC for this service. When Gordon Campbell and Dado Banatao founded Chips and Technologies, which is generally considered the first fabless firm, in 1984, one friend alleged it “wasn’t a real semiconductor company,” since it didn’t build its own chips. However, the graphics chips they designed for PCs proved popular, competing with products built by some of the industry’s biggest players. Eventually Chips and Technologies faded and was purchased by Intel. However, it had proved that a fabless business model could work, requiring only a good idea and a couple of million dollars in startup capital, a tiny fraction of the money needed to build a fab.

对于半导体初创公司来说,计算机图形仍然是一个有吸引力的利基市场,因为与 PC 微处理器不同,英特尔在图形领域并没有事实上的垄断地位。从 IBM 到 Compaq,每家 PC 制造商都必须为其主处理器使用 Intel 或 AMD 芯片,因为这两家公司事实上垄断了 PC 所需的 x86 指令集。在屏幕上渲染图像的芯片市场竞争更加激烈。半导体的出现代工厂,以及启动成本的降低,意味着不仅硅谷贵族可以竞争制造最好的图形处理器。最终主导图形芯片市场的公司,英伟达的起步并不在时尚的帕洛阿尔托咖啡馆,而是在圣何塞粗略地区的丹尼咖啡馆。

Computer graphics remained an appealing niche for semiconductor startups, because unlike PC microprocessors, in graphics Intel didn’t have a de facto monopoly. Every PC maker, from IBM to Compaq, had to use an Intel or an AMD chip for their main processor, because these two firms had a de facto monopoly on the x86 instruction set that PCs required. There was a lot more competition in the market for chips that rendered images on screens. The emergence of semiconductor foundries, and the driving down of startup costs, meant that it wasn’t only Silicon Valley aristocracy that could compete to build the best graphics processors. The company that eventually came to dominate the market for graphics chips, Nvidia, had its humble beginnings not in a trendy Palo Alto coffeehouse but in a Denny’s in a rough part of San Jose.

英伟达由 Chris Malachowsky、Curtis Priem 和 Jensen Huang 于 1993 年创立,后者至今仍担任 CEO。Priem 在 IBM 期间完成了如何计算图形的基础工作,然后与 Malachowsky 一起在 Sun Microsystems 工作。黄原来自台湾,后来搬到肯塔基州小时候,曾在硅谷芯片制造商 LSI 工作。他成为了英伟达的 CEO 和公众形象,总是穿着深色牛仔裤、黑色衬衫和黑色皮夹克,并拥有史蒂夫·乔布斯式的光环,表明他已经深入了解计算的未来。

Nvidia was founded in 1993 by Chris Malachowsky, Curtis Priem, and Jensen Huang, the latter of whom remains CEO today. Priem had done fundamental work on how to compute graphics while at IBM, then worked at Sun Microsystems alongside Malachowsky. Huang, who was originally from Taiwan but had moved to Kentucky as a child, worked for LSI, a Silicon Valley chipmaker. He became the CEO and the public face of Nvidia, always wearing dark jeans, a black shirt, and a black leather jacket, and possessing a Steve Jobs−like aura suggesting that he’d seen far into the future of computing.

Nvidia 的第一批客户——视频和电脑游戏公司——可能看起来不是最前沿的,但该公司认为图形的未来将是生产复杂的 3D 图像。早期的 PC 是一个单调乏味的 2D 世界,因为显示 3D 图像所需的计算量巨大。在 1990 年代,当 Microsoft Office 推出了一款名为 Clippy 的动画回形针,它位于屏幕一侧并提供建议时,它代表了图形技术的飞跃——并且经常导致计算机死机。

Nvidia’s first set of customers—video and computer game companies—might not have seemed like the cutting edge, yet the firm wagered that the future of graphics would be in producing complex, 3D images. Early PCs were a dull, drab, 2D world, because the computation required to display 3D images was immense. In the 1990s, when Microsoft Office introduced an animated, paperclip called Clippy that sat at the side of the screen and dispensed advice, it represented a leap forward in graphics—and often caused computers to freeze.

Nvidia 不仅设计了能够处理 3D 图形的称为图形处理器单元 (GPU) 的芯片,它还围绕它们设计了一个软件生态系统。制作逼真的图形需要使用称为着色器的程序,它告诉图像中的所有像素应该如何在给定的光影中描绘它们。着色器应用于图像中的每个像素,这是对数千个像素进行的相对简单的计算。Nvidia 的 GPU 可以快速渲染图像,因为与 Intel 的微处理器或其他通用 CPU 不同,它们的结构可以同时进行许多简单的计算——比如着色像素——。

Nvidia not only designed chips called graphics processor units (GPUs) capable of handling 3D graphics, it also devised a software ecosystem around them. Making realistic graphics requires use of programs called shaders, which tell all the pixels in an image how they should be portrayed in, say, a given shade of light. The shader is applied to each of the pixels in an image, a relatively straightforward calculation conducted over many thousands of pixels. Nvidia’s GPUs can render images quickly because, unlike Intel’s microprocessors or other general-purpose CPUs, they’re structured to conduct lots of simple calculations—like shading pixels—simultaneously.

2006 年,Nvidia 意识到高速并行计算可以用于计算机图形以外的用途,因此发布了 CUDA,该软件允许使用标准编程语言对 GPU 进行编程,而无需任何图形参考。即使 Nvidia 正在生产一流的图形芯片,黄据一家公司在 2017 年的估计,在这项软件工作上花费了至少 100 亿美元,让任何程序员——不仅仅是图形专家——都可以使用 Nvidia 的芯片。Huang 免费赠送了 CUDA,但该软件仅适用于 Nvidia 的芯片。通过使芯片在图形行业之外有用,Nvidia发现了一个巨大的并行处理新市场,从计算化学到天气预报。当时,黄只能模糊地感知到并行处理的最大用例的潜在增长:人工智能。

In 2006, realizing that high-speed parallel computations could be used for purposes besides computer graphics, Nvidia released CUDA, software that lets GPUs be programmed in a standard programming language, without any reference to graphics at all. Even as Nvidia was churning out top-notch graphics chips, Huang spent lavishly on this software effort, at least $10 billion, according to a company estimate in 2017, to let any programmer—not just graphics experts—work with Nvidia’s chips. Huang gave away CUDA for free, but the software only works with Nvidia’s chips. By making the chips useful beyond the graphics industry, Nvidia discovered a vast new market for parallel processing, from computational chemistry to weather forecasting. At the time, Huang could only dimly perceive the potential growth in what would become the biggest use case for parallel processing: artificial intelligence.

今天,Nvidia 的芯片主要由台积电制造,在最先进的数据中心中都可以找到。公司不需要建立自己的晶圆厂,这是一件好事。在启动阶段,可能不可能筹集到必要的资金。给在 Denny's 工作的芯片设计师几百万美元已经是一场赌博。下注超过一亿美元——当时新晶圆厂的成本——即使对于硅谷最具冒险精神的投资者来说也是一项艰巨的任务。此外,正如杰里·桑德斯(Jerry Sanders)所指出的,运营一个晶圆厂既昂贵又耗时。像英伟达那样简单地设计一流的芯片已经够难的了。如果它还必须管理自己的制造流程,它可能就没有资源或带宽来投入资金来构建软件生态系统。

Today Nvidia’s chips, largely manufactured by TSMC, are found in most advanced data centers. It’s a good thing the company didn’t need to build its own fab. At the startup stage, it would probably have been impossible to raise the necessary sums. Giving a couple million dollars to chip designers working in a Denny’s was already a gamble. Betting over a hundred million dollars—the cost of a new fab at the time—would have been a stretch even for Silicon Valley’s most adventurous investors. Moreover, as Jerry Sanders noted, running a fab well is expensive and time-consuming. It’s hard enough simply to design top-notch chips, as Nvidia did. If it had also had to manage its own manufacturing processes, it probably wouldn’t have had the resources or the bandwidth to plow money into building a software ecosystem.

Nvidia 并不是唯一一家为专用逻辑芯片开创新用例的无晶圆厂公司。通信理论教授欧文·雅各布斯(Irwin Jacobs)高举微处理器并宣称“这就是未来!” 在 1970 年代初期的一次学术会议上,现在相信未来已经到来。移动电话——附着在汽车仪表板或地板上的大型黑色塑料砖——即将进入第二代 (2G) 技术。电话公司是试图就让他们的手机相互通信的技术标准达成一致。大多数公司都想要一个称为“时分多址”的系统,这样来自多个电话的数据将在相同的无线电波频率上传输,当有片刻沉默时,来自一个电话的数据会被插入无线电波频谱在不同的通话中。

Nvidia wasn’t the only fabless company pioneering new use cases for specialized logic chips. Irwin Jacobs, the communications theory professor who’d held aloft a microprocessor and declared “This is the future!” at an academic conference in the early 1970s, now believed the future had arrived. Mobile phones—big, black bricks of plastic that were attached to the dashboard or floor of a car—were about to enter their second generation (2G) of technology. Phone companies were trying to agree on a technology standard that would let their phones communicate with one other. Most companies wanted a system called “time-division multiple access,” whereby data from multiple phone calls would be transmitted on the same radio-wave frequency, with data from one call slotted into the radio-wave spectrum when there was a moment of silence in a different call.

Jacobs 一如既往地坚信摩尔定律,他认为更复杂的跳频系统会更好地工作。他建议在不同频率之间移动通话数据,而不是将给定电话保持在某个频率上,让他将更多电话塞进可用的频谱空间。大多数人认为他在理论上是对的,但这样的系统在实践中永远行不通。他们争辩说,语音质量会很低,电话也会被挂断。在频率之间移动呼叫数据并让另一端的电话对其进行解释所需的处理量似乎是巨大的。

Jacobs, whose faith in Moore’s Law was as strong as ever, thought a more complicated system of frequency-hopping would work better. Rather than keeping a given phone call on a certain frequency, he proposed moving call data between different frequencies, letting him cram more calls into available spectrum space. Most people thought he was right in theory, but that such a system would never work in practice. Voice quality would be low, they argued, and calls would be dropped. The amount of processing needed to move call data between frequencies and have it interpreted by a phone on the other end seemed enormous.

雅各布斯不同意,他在 1985 年创立了一家名为高通(Quality Communications)的公司来证明这一点。他建立了一个带有几个蜂窝塔的小型网络,以证明它可以工作。很快,整个行业都意识到,高通的系统可以通过依靠摩尔定律运行能够理解所有反弹的无线电波的算法,将更多的手机通话容纳到现有的频谱空间中。

Jacobs disagreed, founding a company called Qualcomm—Quality Communications—in 1985 to prove the point. He built a small network with a couple cell towers to prove it would work. Soon the entire industry realized Qualcomm’s system would make it possible to fit far more cell phone calls into existing spectrum space by relying on Moore’s Law to run the algorithms that make sense of all the radio waves bouncing around.

对于 2G 之后的每一代手机技术,高通都贡献了关于如何通过无线电频谱传输更多数据的关键思想,并出售了具有能够破译这种杂音信号的计算能力的专用芯片。该公司的专利非常重要没有它们就不可能制造手机。高通很快多元化进入新的业务线,不仅设计手机中与蜂窝网络通信的调制解调器芯片,还设计运行智能手机核心系统的应用处理器。这些芯片设计都是巨大的工程成就,每一个都建立在数千万行代码。高通通过销售芯片和授权知识赚了数千亿美元财产。但它没有制造任何芯片:它们都是内部设计的,但由三星或台积电等公司制造。

For each generation of cell phone technology after 2G, Qualcomm contributed key ideas about how to transmit more data via the radio spectrum and sold specialized chips with the computing power capable of deciphering this cacophony of signals. The company’s patents are so fundamental it’s impossible to make a cell phone without them. Qualcomm soon diversified into a new business line, designing not only the modem chips in a phone that communicate with a cell network, but also the application processors that run a smartphone’s core systems. These chip designs are monumental engineering accomplishments, each built on tens of millions of lines of code. Qualcomm has made hundreds of billions of dollars selling chips and licensing intellectual property. But it hasn’t fabricated any chips: they’re all designed in-house but fabricated by companies like Samsung or TSMC.

很容易感叹半导体制造的离岸外包。但如果像高通这样的公司不得不每年投资数十亿美元建造晶圆厂,它们可能无法生存。雅各布斯和他的工程师是把数据塞进无线电波频谱并设计更聪明的芯片来解码这些信号的含义的奇才。与 Nvidia 的情况一样,他们也不必尝试成为半导体制造专家,这是一件好事。高通一再考虑开设自己的制造设施,但考虑到所涉及的成本和复杂性,始终拒绝这样做。感谢台积电、三星和其他愿意生产芯片的公司,高通的工程师可以专注于他们在频谱管理和半导体设计方面的核心优势。

It’s easy to lament the offshoring of semiconductor manufacturing. But companies like Qualcomm might not have survived if they’d had to invest billions of dollars each year building fabs. Jacobs and his engineers were wizards at cramming data into the radio-wave spectrum and devising ever-more-clever chips to decode the meaning of these signals. As was the case with Nvidia, it was a good thing they didn’t have to try to be semiconductor manufacturing experts, too. Qualcomm repeatedly considered opening its own fabrication facilities, but always decided against it, given the cost and complexity involved. Thanks to TSMC, Samsung, and other companies willing to produce their chips, Qualcomm’s engineers could focus on their core strengths in managing spectrum and in semiconductor design.

还有许多其他美国芯片公司从无晶圆厂模式中受益,让他们无需花费数十亿美元建立内部晶圆厂就可以生产新的芯片设计。出现了整个新类别的芯片,这些芯片仅在台积电和其他代工厂而非内部制造。现场可编程门阵列是一种可以针对不同用途进行编程的芯片,由 Xilinx 和 Altera 等公司开创,这两家公司从一开始就依赖外包制造。然而,最大的变化不仅仅是新型芯片。通过使手机、高级图形和并行处理成为可能,无晶圆厂公司启用了全新类型的计算。

There were many other U.S. chip firms that benefitted from a fabless model, letting them produce new chip designs without having to spend billions building an in-house fab. Entire new categories of chips emerged that were fabricated only at TSMC and other foundries rather than in-house. Field-programmable gate arrays, chips that can be programmed for different uses, were pioneered by companies like Xilinx and Altera, both of which relied on outsourced manufacturing from their earliest days. The biggest change, however, wasn’t simply new types of chips. By making possible mobile phones, advanced graphics, and parallel processing, fabless firms enabled entirely new types of computing.

第 37 章 张忠谋的大同盟

CHAPTER 37 Morris Chang’s Grand Alliance

Ĵerry Sanders 可能已经承诺永远不会放弃他的晶圆厂,但是用小刀和镊子设计芯片的那一代工程师正在离开现场。他们的接班人接受过计算机科学新学科的培训,许多人主要通过 1980 年代和 1990 年代出现的新芯片设计软件程序了解半导体。对硅谷的许多人来说,桑德斯对晶圆厂的浪漫依恋就像他的大男子主义招摇一样格格不入。在 2000 年代和 2010 年代接管美国半导体公司的新 CEO 阶层倾向于使用 MBA 和博士的语言,在季度财报电话会议上与华尔街分析师随意谈论资本支出和利润率。以大多数标准衡量,新一代的高管人才比创建硅谷的化学家和物理学家专业得多。但与之前的巨人相比,它们往往显得陈旧。

Jerry Sanders may have promised never to give up his fabs, but the generation of engineers who came of age designing chips with penknives and tweezers was leaving the scene. Their replacements had been trained in the new discipline of computer science, and many knew semiconductors primarily through the new chip design software programs that emerged out of the 1980s and 1990s. To many people in Silicon Valley, Sanders’s romantic attachment to fabs seemed as out of touch as his macho swagger. The new class of CEOs who took over America’s semiconductor firms in the 2000s and 2010s tended to speak the language of MBAs as well as PhDs, chatting casually about capex and margins with Wall Street analysts on quarterly earnings calls. By most measures this new generation of executive talent was far more professional than the chemists and physicists who’d built Silicon Valley. But they often seemed stale in comparison to the giants who preceded them.

一个对不可能的技术疯狂下注的时代正在被更有组织、更专业和更合理化的东西所取代。豪赌取而代之的是有计划的风险管理。在这个过程中很难摆脱丢失了什么的感觉。在芯片行业的创始人中,只有张忠谋留下了,他在台湾的办公室里抽着烟斗,他为 216his 的健康辩护,这是一种有益于健康的习惯,或者至少是为了他的心情。在 2000 年代,甚至 Chang 也开始考虑继任计划。2005 年,74 岁的他辞去了 CEO 一职,但仍担任台积电董事长。很快就没有人记得在实验室里和杰克·基尔比一起工作,或者和鲍勃·诺伊斯一起喝过啤酒。

An era of wild wagers on impossible technologies was being superseded by something more organized, professionalized, and rationalized. Bet-the-house gambles were replaced by calculated risk management. It was hard to escape the sense that something was lost in the process. Of the chip industry’s founders, only Morris Chang remained, smoking his pipe in his office in Taiwan, a habit he defended as good for 216his health, or at least for his mood. In the 2000s, even Chang began to think about succession planning. In 2005, aged seventy-four, he stepped down from the role of CEO, though he remained chairman of TSMC. Soon there’d be no one left who remembered working in the lab alongside Jack Kilby or drinking beers with Bob Noyce.

芯片行业高层的换岗加速了芯片设计和制造的分裂,后者大部分都离岸。桑德斯从 AMD 退休五年后,该公司宣布将拆分其芯片设计和制造业务。华尔街欢呼雀跃,认为如果没有资本密集型晶圆厂,新的 AMD 将更有利可图。AMD 将这些设施分拆成一家新公司,该公司将像台积电一样作为代工厂运营,不仅为 AMD 生产芯片,还为其他客户生产芯片。阿布扎比政府的投资部门穆巴达拉成为新铸造厂的主要投资者,对于一个以碳氢化合物而非高科技着称的国家来说,这是一个出人意料的位置。审查外国战略资产购买情况的美国政府机构 CFIUS 认为这不会对国家安全造成影响,因此放弃了此次出售。但 AMD 生产能力的命运最终将塑造芯片行业——并保证最先进的芯片制造将在海外进行。

The changing of the guard atop the chip industry accelerated the splitting of chip design and manufacturing, with much of the latter offshored. Five years after Sanders retired from AMD, the company announced it was dividing its chip design and fabrication businesses. Wall Street cheered, reckoning the new AMD would be more profitable without the capital-intensive fabs. AMD spun out these facilities into a new company that would operate as a foundry like TSMC, producing chips not only for AMD but other customers, too. The investment arm of the Abu Dhabi government, Mubadala, became the primary investor in the new foundry, an unexpected position for a country known more for hydrocarbons than for high-tech. CFIUS, the U.S. government body that reviews foreign purchases of strategic assets, waved the sale through, judging that it had no national security implications. But the fate of AMD’s production capabilities would end up shaping the chip industry—and guaranteeing that the most advanced chipmaking would take place offshore.

GlobalFoundries,作为这家继承 AMD 晶圆厂的新公司而闻名,进入了一个与以往一样竞争激烈且无情的行业。摩尔定律在 2000 年代和 2010 年代初向前推进,迫使尖端芯片制造商花费越来越大的资金推出一种新的、更先进的制造工艺,大约每两年一次。智能手机、PC 和服务器芯片迅速迁移到每个新的“节点”,利用更高的处理能力和更低的功耗,因为晶体管更加密集。每个节点转换都需要更昂贵的机器来生产。

GlobalFoundries, as this new company that inherited AMD’s fabs was known, entered an industry that was as competitive and unforgiving as ever. Moore’s Law marched forward through the 2000s and early 2010s, forcing cutting-edge chipmakers to spend ever larger sums rolling out a new, more advanced manufacturing process roughly once every two years. Smartphone, PC, and server chips quickly migrated to each new “node,” taking advantage of increased processing power and lower power consumption as transistors were more densely packed. Each node transition required ever more expensive machinery to produce.

多年来,每一代制造技术都以晶体管的栅极长度命名,硅的一部分芯片的导电性将被打开和关闭,从而创建和中断电路。180nm 节点于 1999 年率先推出,随后是 130nm、90nm、65nm 和 45nm,每一代晶体管都缩小到足以在同一区域内塞进大约两倍的晶体管。这降低了每个晶体管的功耗,因为较小的晶体管需要较少的电子流过它们。

For many years, each generation of manufacturing technology was named after the length of the transistor’s gate, the part of the silicon chip whose conductivity would be turned on and off, creating and interrupting the circuit. The 180nm node was pioneered in 1999, followed by 130nm, 90nm, 65nm, and 45nm, with each generation shrinking transistors enough to make it possible to cram roughly twice as many in the same area. This reduced power consumption per transistor, because smaller transistors needed fewer electrons to flow through them.

大约在 2010 年代初,通过二维收缩来更密集地封装晶体管变得不可行。一个挑战是,随着晶体管根据摩尔定律缩小,导体通道的狭窄长度偶尔会导致电源“泄漏”通过电路,即使在开关关闭时也是如此。最重要的是,每个晶体管顶部的二氧化硅层变得非常薄,以至于像“隧道效应”这样的量子效应——跨越经典物理学认为应该无法克服的障碍——开始严重影响晶体管的性能。到 2000 年代中期,每个晶体管顶部的二氧化硅层只有几个原子厚,太小而无法覆盖硅中的所有电子。

Around the early 2010s, it became unfeasible to pack transistors more densely by shrinking them two dimensionally. One challenge was that, as transistors were shrunk according to Moore’s Law, the narrow length of the conductor channel occasionally caused power to “leak” through the circuit even when the switch was off. On top of this, the layer of silicon dioxide atop each transistor became so thin that quantum effects like “tunneling”—jumping through barriers that classical physics said should be insurmountable—began seriously impacting transistor performance. By the mid-2000s, the layer of silicon dioxide on top of each transistor was only a couple of atoms thick, too small to keep a lid on all the electrons sitting in the silicon.

为了更好地控制电子的运动,需要新材料和晶体管设计。与自 1960 年代以来使用的 2D 设计不同,22nm 节点引入了一种新的 3D 晶体管,称为 FinFET(发音为 finfet),它将电路的两端和连接它们的半导体材料通道设置在块的顶部,看起来就像鲸鱼背上伸出的鳍一样。因此,连接电路两端的通道不仅可以从鳍片的顶部施加电场,还可以从鳍片的侧面施加电场,从而增强对电子的控制并克服威胁新一代性能的漏电问题。微小的晶体管。这些纳米级 3D 结构对于摩尔定律的生存至关重要,但它们的制造难度惊人,在沉积、蚀刻和光刻方面需要更高的精度。这增加了主要芯片制造商是否都会完美地执行向 FinFET 架构的转换,或者是否会落后的不确定性。

To better control the movement of electrons, new materials and transistor designs were needed. Unlike the 2D design used since the 1960s, the 22nm node introduced a new 3D transistor, called a FinFET (pronounced finfet), that sets the two ends of the circuit and the channel of semiconductor material that connects them on top of a block, looking like a fin protruding from a whale’s back. The channel that connects the two ends of the circuit can therefore have an electric field applied not only from the top but also from the sides of the fin, enhancing control over the electrons and overcoming the electricity leakage that was threatening the performance of new generations of tiny transistors. These nanometer-scale 3D structures were crucial for the survival of Moore’s Law, but they were staggeringly difficult to make, requiring even more precision in deposition, etching, and lithography. This added uncertainty about whether the major chipmakers would all flawlessly execute the switch to FinFET architectures or whether one might fall behind.

当 GlobalFoundries 于 2009 年作为一家独立公司成立时,行业分析师认为它在这场向 3D 晶体管的竞争中赢得了市场份额。该公司的前高管承认,即使是台积电也很担心。GlobalFoundries 继承了德国的一座大型晶圆厂,并正在纽约建造一座新的尖端工厂。与其竞争对手不同,它将把其最先进的生产能力建立在发达经济体,而不是亚洲。该公司与 IBM 和三星建立了合作伙伴关系,共同开发技术,使客户可以直接与 GlobalFoundries 或三星签订合同来生产他们的芯片。此外,无晶圆厂芯片设计公司渴望有一个可靠的竞争对手台积电,因为台湾的庞然大物已经拥有全球大约一半的代工市场。

When GlobalFoundries was established as an independent company in 2009, industry analysts thought it was well placed to win market share amid this race toward 3D transistors. Even TSMC was worried, the company’s former executives admit. GlobalFoundries had inherited a massive fab in Germany and was building a new, cutting-edge facility in New York. Unlike its rivals, it would be basing its most advanced production capacity in advanced economies, not in Asia. The company had a partnership with IBM and Samsung to jointly develop technology, making it straightforward for customers to contract with either GlobalFoundries or with Samsung to produce their chips. Moreover, fabless chip design firms were hungry for a credible competitor to TSMC, because the Taiwanese behemoth already had around half of the world’s foundry market.

唯一的另一个主要竞争对手是三星,其代工业务的技术与台积电大致相当,尽管该公司的生产能力要少得多。不过,由于三星的部分业务涉及构建其内部设计的芯片,因此出现了复杂情况。像台积电这样的公司为数十家客户制造芯片并不懈地致力于让他们满意,而三星拥有自己的智能手机和其他消费电子产品线,因此它正在与许多客户竞争。这些公司担心与三星芯片代工厂共享的想法可能最终会出现在其他三星产品中。台积电和 GlobalFoundries 没有此类利益冲突。

The only other major competitor was Samsung, whose foundry business had technology that was roughly comparable to TSMC’s, though the company possessed far less production capacity. Complications arose, though, because part of Samsung’s operation involved building chips that it designed in-house. Whereas a company like TSMC builds chips for dozens of customers and focuses relentlessly on keeping them happy, Samsung had its own line of smartphones and other consumer electronics, so it was competing with many of its customers. Those firms worried that ideas shared with Samsung’s chip foundry might end up in other Samsung products. TSMC and GlobalFoundries had no such conflicts of interest.

转向 FinFET 晶体管并不是在 GlobalFoundries 成立之际对芯片行业造成的唯一冲击。台积电的 40nm 工艺面临着巨大的制造问题,给 GlobalFoundries有机会将自己与大型竞争对手区分开来。此外,2008-2009 年的金融危机威胁着芯片行业的重组。消费者停止购买电子产品,因此科技公司停止订购芯片。半导体采购下滑。它台积电的一位高管回忆说,感觉就像一部电梯从空荡荡的竖井中倾泻而下。如果有什么事情可以扰乱芯片行业,那就是全球金融危机。

The move to FinFET transistors wasn’t the only shock to the chip industry that coincided with the establishment of GlobalFoundries. TSMC faced substantial manufacturing problems with its 40nm process, giving GlobalFoundries a chance to distinguish itself from its large rival. Moreover, the 2008−2009 financial crisis was threatening to reorder the chip industry. Consumers stopped buying electronics, so tech firms stopped ordering chips. Semiconductor purchases slumped. It felt like an elevator careening down an empty shaft, one TSMC executive recalled. If anything could disrupt the chip industry, a global financial crisis was it.

不过,Morris Chang 并不打算放弃代工业务的主导地位。自从他的老同事杰克·基尔比发明集成电路以来,他经历了每一个行业周期。他确信这种低迷最终也会结束。过度扩张的公司将被淘汰出局,而那些在经济低迷时期进行投资的公司则处于抢占市场份额的位置。此外,Chang 很早就意识到智能手机将如何改变计算方式,从而也将如何改变芯片行业。媒体关注的是像 Facebook 的马克·扎克伯格这样的年轻科技大亨,但 77 岁的张有一个很少有人能比得上的观点。他告诉福布斯,移动设备将成为芯片行业的“游戏规则改变者”,认为它们预示着与 PC 带来的重大转变。他致力于赢得这项业务的最大份额,不惜一切代价。

Morris Chang wasn’t about to give up dominance of the foundry business, though. He’d lived through every industry cycle since his old colleague Jack Kilby invented the integrated circuit. He was sure this downturn would eventually end, too. Companies that were overextended would be pushed out of business, leaving those that invested during the downturn positioned to grab market share. Moreover, Chang realized as early as anyone how smartphones would transform computing—and therefore how they would change the chip industry, too. The media focused on young tech tycoons like Facebook’s Mark Zuckerberg, but seventy-seven-year-old Chang had a perspective that few could match. Mobile devices would be a “game-changer” for the chip industry, he told Forbes, perceiving them as heralding shifts as significant as the PC had brought. He was committed to winning the lion’s share of this business, whatever the cost.

Chang意识到台积电可以在技术上领先竞争对手,因为它是其他公司设计产品的中立参与者。他将此称为台积电的“大联盟”,由数十家设计芯片、销售知识产权、生产材料或制造机械的公司组成。这些公司中有许多相互竞争,但由于没有一家制造晶圆,因此没有一家与台积电竞争。因此,台积电可以在它们之间进行协调,制定芯片行业大多数其他公司都同意使用的标准。他们别无选择,因为与台积电工艺的兼容性对几乎每家公司都至关重要。对于无晶圆厂公司来说,台积电是他们最具竞争力的制造服务来源。对于设备公司和材料公司来说,台积电往往是他们最大的客户。随着智能手机开始起飞,推动了对硅的需求,Morris Chang 坐在了中心位置。“台积电知道利用每个人的创新很重要,”Chang 宣称,“我们、设备制造商、客户和 IP 提供商的创新。这就是大联盟的力量。” 这对财务的影响是深远的。“台积电及其十大客户的研发支出加起来,”他吹嘘道,“超过了三星和英特尔在一起。” 旧的设计与制造一体化的模式在其他行业处于竞争状态时将难以竞争围绕台积电合并。

Chang realized that TSMC could pull ahead of rivals technologically because it was a neutral player around which other companies would design their products. He called this TSMC’s “Grand Alliance,” a partnership of dozens of companies that design chips, sell intellectual property, produce materials, or manufacture machinery. Many of these companies compete with each other, but since none fabricate wafers, none compete with TSMC. TSMC could therefore coordinate between them, setting standards that most other companies in the chip industry would agree to use. They had no choice, because compatibility with TSMC’s processes was crucial for almost every company. For fabless firms, TSMC was their most competitive source of manufacturing services. For equipment companies and materials firms, TSMC was often their biggest customer. As smartphones began to take off, driving up demand for silicon, Morris Chang sat at the center. “TSMC knows it is important to use everyone’s innovation,” Chang declared, “ours, that of the equipment makers, of our customers, and of the IP providers. That’s the power of the Grand Alliance.” The financial implications of this were profound. “The combined R&D spending of TSMC and its ten biggest customers,” he bragged “exceeds that of Samsung and Intel together.” The old model of integrating design and manufacture would struggle to compete when the rest of the industry was coalescing around TSMC.

台积电在半导体领域的中心地位要求它有能力为所有最大的客户生产芯片。这样做并不便宜。在金融危机期间,Chang 精心挑选的继任者 Rick Tsai 做了几乎所有 CEO 都会做的事情——裁员和削减成本。张想反其道而行之。让公司的 40nm 芯片制造重回正轨需要对人员和技术进行投资。为了赢得更多的智能手机业务——尤其是 2007 年推出的苹果 iPhone,它最初从台积电的主要竞争对手三星那里购买了关键芯片——需要对芯片制造能力进行大量投资。张认为蔡的削减成本是失败主义者。“投资非常非常少,”张后来告诉记者。“我一直认为公司有更多的能力...... 它没有发生。出现了停滞。”

TSMC’s position at the center of the semiconductor universe required it to have capacity to produce chips for all its biggest customers. Doing so wouldn’t be cheap. Amid the financial crisis, Chang’s handpicked successor, Rick Tsai, had done what nearly every CEO did—lay off employees and cut costs. Chang wanted to do the opposite. Getting the company’s 40nm chipmaking back on track required investing in personnel and technology. Trying to win more smartphone business—especially that of Apple’s iPhone, which launched in 2007 and which initially bought its key chips from TSMC’s archrival, Samsung—required massive investment in chipmaking capacity. Chang saw Tsai’s cost cutting as defeatist. “There was very, very little investment,” Chang told journalists afterward. “I had always thought that the company was capable of more…. It didn’t happen. There was stagnation.”

于是张开除继任者,重新控制台积电。该公司股价当天下跌,因为投资者担心他会推出一个回报不确定的风险支出计划。张认为真正的风险是接受现状。他不打算让金融危机威胁台积电争夺行业领导地位。他在芯片制造方面拥有长达半个世纪的记录,自 1950 年代中期以来,他一直在磨练这一声誉。因此,在危机最严重的时候,Chang 重新雇用了前任 CEO 解雇的员工,并在新产能和研发方面的投资翻了一番。尽管发生了危机,但他宣布在 2009 年和 2010 年增加了数十亿美元的资本支出。最好“有容量比相反的要大,”Chang 宣称。任何想打入代工业务的人都将面临台积电的全面竞争,因为它正在竞相占领蓬勃发展的智能手机芯片市场。“我们才刚刚开始,”Chang 在 2012 年宣称,他在半导体行业进入了他的第六个十年。

So Chang fired his successor and retook direct control of TSMC. The company’s stock price fell that day, as investors worried he’d launch a risky spending program with uncertain returns. Chang thought the real risk was accepting the status quo. He wasn’t about to let a financial crisis threaten TSMC in the race for industry leadership. He had a half-century-long track record at chipmaking, a reputation he’d honed since the mid-1950s. So at the depths of the crisis Chang rehired the workers the former CEO had laid off and doubled down on investment in new capacity and R&D. He announced several multibillion-dollar increases to capital spending in 2009 and 2010 despite the crisis. It was better “to have too much capacity than the other way around,” Chang declared. Anyone who wanted to break into the foundry business would face the full force of competition from TSMC as it raced to capture the booming market for smartphone chips. “We’re just at the start,” Chang declared in 2012, as he launched into his sixth decade atop the semiconductor industry.

第 38 章 苹果硅

CHAPTER 38 Apple Silicon

台积电等代工厂崛起的最大受益者是一家大多数人甚至没有意识到设计芯片的公司:苹果。然而,史蒂夫乔布斯建立的公司一直专注于硬件,因此苹果完善其设备的愿望包括控制内部的硅也就不足为奇了。从他在苹果公司的早期开始,史蒂夫乔布斯就深入思考了软件和硬件之间的关系。1980 年,当他的头发几乎长到肩膀,小胡子遮住上唇时,乔布斯做了一次演讲,他问道:“什么是软件?” “我唯一能想到的,”他回答说,“软件是一种变化太快的东西,或者你还不完全知道你想要什么,或者你没有时间把它变成硬件。”

The greatest beneficiary of the rise of foundries like TSMC was a company that most people don’t even realize designs chips: Apple. The company Steve Jobs built has always specialized in hardware, however, so it’s no surprise that Apple’s desire to perfect its devices includes controlling the silicon inside. Since his earliest days at Apple, Steve Jobs had thought deeply about the relationship between software and hardware. In 1980, when his hair nearly reached his shoulders and his mustache covered his upper lip, Jobs gave a lecture that asked, “What is software?” “The only thing I can think of,” he answered, “is software is something that is changing too rapidly, or you don’t exactly know what you want yet, or you didn’t have time to get it into hardware.”

乔布斯没有时间将他所有的想法都投入到第一代 iPhone 的硬件中,第一代 iPhone 使用的是苹果自己的 iOS 操作系统,但将其芯片的设计和生产外包给了三星。这款革命性的新手机还有许多其他芯片:英特尔内存芯片、Wolfson 设计的音频处理器、德国英飞凌生产的连接蜂窝网络的调制解调器、CSR 设计的蓝牙芯片和 Skyworks 的信号放大器,其中。都是由其他公司设计的。

Jobs didn’t have time to get all his ideas into the hardware of the first-generation iPhone, which used Apple’s own iOS operating system but outsourced design and production of its chips to Samsung. The revolutionary new phone had many other chips, too: an Intel memory chip, an audio processor designed by Wolfson, a modem to connect with the cell network produced by Germany’s Infineon, a Bluetooth chip designed by CSR, and a signal amplifier from Skyworks, among others. All were designed by other companies.

随着乔布斯推出新版本的 iPhone,他开始将他对智能手机的愿景刻入苹果自己的硅芯片中。在推出 iPhone 一年后,Apple 收购了一家名为 PA Semi 的小型硅谷芯片设计公司,该公司在节能处理方面拥有专长。很快,Apple 开始招聘一些业内最优秀的芯片设计师。两年后,该公司宣布设计了自己的应用处理器 A4,并用于新 iPad 和 iPhone 4。设计与运行智能手机的处理器一样复杂的芯片非常昂贵,这就是大多数中低端智能手机公司从高通等公司购买现成芯片的原因。然而,苹果在巴伐利亚和以色列以及硅谷的研发和芯片设计设施上投入了大量资金,工程师们在那里设计了最新的芯片。现在,Apple 不仅为其大多数设备设计主处理器,还设计了运行 AirPods 等配件的辅助芯片。这项对专用硅的投资解释了为什么苹果的产品运行如此顺畅。在 iPhone 推出后的四年内,苹果公司正在改头换面全球 60% 的利润来自智能手机销售,击败了诺基亚和黑莓等竞争对手,让东亚智能手机制造商在低利润市场竞争廉价手机。

As Jobs introduced new versions of the iPhone, he began etching his vision for the smartphone into Apple’s own silicon chips. A year after launching the iPhone, Apple bought a small Silicon Valley chip design firm called PA Semi that had expertise in energy-efficient processing. Soon Apple began hiring some of the industry’s best chip designers. Two years later, the company announced it had designed its own application processor, the A4, which it used in the new iPad and the iPhone 4. Designing chips as complex as the processors that run smartphones is expensive, which is why most low- and midrange smartphone companies buy off-the-shelf chips from companies like Qualcomm. However, Apple has invested heavily in R&D and chip design facilities in Bavaria and Israel as well as Silicon Valley, where engineers design its newest chips. Now Apple not only designs the main processors for most of its devices but also ancillary chips that run accessories like AirPods. This investment in specialized silicon explains why Apple’s products work so smoothly. Within four years of the iPhone’s launch, Apple was making over 60 percent of all the world’s profits from smartphone sales, crushing rivals like Nokia and BlackBerry and leaving East Asian smartphone makers to compete in the low-margin market for cheap phones.

就像高通和其他推动移动革命的芯片公司一样,尽管苹果设计了更多的芯片,但它并没有制造任何这些芯片。苹果以将其手机、平板电脑和其他设备的组装外包给数十万客户而闻名中国的装配线工人,他们负责将小零件拧紧和粘合在一起。中国的组装设施生态系统是世界上构建电子设备的最佳场所。台湾公司,如富士康和纬创资通,在中国为苹果公司运营这些设施,具有生产手机、个人电脑和其他电子产品的独特能力。尽管东莞和郑州等中国城市的电子组装设施是世界上效率最高的,但它们并非不可替代。世界上仍有数亿自给自足的农民,他们乐于将组件固定在 iPhone 上每小时一美元。富士康在中国组装了大部分苹果产品,但也有一些是在中国制造的越南和印度也是。

Like Qualcomm and the other chip firms that powered the mobile revolution, even though Apple designs ever more silicon, it doesn’t build any of these chips. Apple is well known for outsourcing assembly of its phones, tablets, and other devices to several hundred thousand assembly line workers in China, who are responsible for screwing and gluing tiny pieces together. China’s ecosystem of assembly facilities is the world’s best place to build electronic devices. Taiwanese companies, like Foxconn and Wistron, that run these facilities for Apple in China are uniquely capable of churning out phones, PCs, and other electronics. Though the electronics assembly facilities in Chinese cities like Dongguan and Zhengzhou are the world’s most efficient, however, they aren’t irreplaceable. The world still has several hundred million subsistence farmers who’d happily fasten components into an iPhone for a dollar an hour. Foxconn assembles most of its Apple products in China, but it builds some in Vietnam and India, too.

与流水线工人不同,智能手机内部的芯片很难更换。随着晶体管的缩小,它们变得越来越难以制造。能够制造尖端芯片的半导体公司数量已经减少。到 2010 年,在苹果推出第一款芯片的时候,只有少数几家尖端代工厂:台湾的台积电、韩国的三星,以及——也许还有——格罗方德,这取决于它能否成功赢得市场份额。英特尔仍然是缩小晶体管的世界领先者,它仍然专注于为个人电脑和服务器制造自己的芯片,而不是为其他公司的手机制造处理器。像中芯国际这样的中国代工厂正在努力追赶,但仍落后多年。

Unlike assembly line workers, the chips inside smartphones are very difficult to replace. As transistors have shrunk, they’ve become ever harder to fabricate. The number of semiconductor companies that can build leading-edge chips has dwindled. By 2010, at the time Apple launched its first chip, there were just a handful of cutting-edge foundries: Taiwan’s TSMC, South Korea’s Samsung, and—perhaps—GlobalFoundries, depending on whether it could succeed in winning market share. Intel, still the world’s leader at shrinking transistors, remained focused on building its own chips for PCs and servers rather than processors for other companies’ phones. Chinese foundries like SMIC were trying to catch up but remained years behind.

正因为如此,智能手机供应链看起来与与个人电脑相关的供应链非常不同。智能手机和个人电脑都主要在中国组装,高价值组件主要在美国、欧洲、日本或韩国设计。对于个人电脑,大多数处理器来自英特尔,并在该公司位于美国、爱尔兰或以色列的工厂之一生产。智能手机是不同的。它们塞满了芯片,不仅是主处理器(苹果自己设计的),还有用于连接蜂窝网络的调制解调器和射频芯片、用于 WiFi 和蓝牙连接的芯片、用于相机的图像传感器、至少两个内存芯片、感应运动的芯片(因此您的手机知道何时将其水平转动),以及管理电池、音频和无线充电的半导体。

Because of this, the smartphone supply chain looks very different from the one associated with PCs. Smartphones and PCs are both assembled largely in China with high-value components mostly designed in the U.S., Europe, Japan, or Korea. For PCs, most processors come from Intel and are produced at one of the company’s fabs in the U.S., Ireland, or Israel. Smartphones are different. They’re stuffed full of chips, not only the main processor (which Apple designs itself), but modem and radio frequency chips for connecting with cellular networks, chips for WiFi and Bluetooth connections, an image sensor for the camera, at least two memory chips, chips that sense motion (so your phone knows when you turn it horizontal), as well as semiconductors that manage the battery, the audio, and wireless charging. These chips make up most of the bill of materials needed to build a smartphone.

随着半导体制造能力转移到台湾和韩国,生产许多此类芯片的能力也随之转移。应用处理器,即每部智能手机内部的电子大脑,主要在台湾和韩国生产,然后运往中国,在手机的塑料外壳和玻璃屏幕内进行最终组装。苹果的 iPhone 处理器是在台湾独家制造的。今天,除了台积电,没有任何一家公司有能力或生产能力制造苹果需要的芯片。因此,每部 iPhone 的背面都刻有文字——“由加利福尼亚的 Apple 设计。在中国组装”——极具误导性。iPhone最不可替代的部件确实是在加州设计,在中国组装的。但它们只能在台湾制造。

As semiconductor fabrication capacity migrated to Taiwan and South Korea, so too did the ability to produce many of these chips. Application processors, the electronic brain inside each smartphone, are mostly produced in Taiwan and South Korea before being sent to China for final assembly inside a phone’s plastic case and glass screen. Apple’s iPhone processors are fabricated exclusively in Taiwan. Today, no company besides TSMC has the skill or the production capacity to build the chips Apple needs. So the text etched onto the back of each iPhone—“Designed by Apple in California. Assembled in China”—is highly misleading. The iPhone’s most irreplaceable components are indeed designed in California and assembled in China. But they can only be made in Taiwan.

第 39 章 极紫外

CHAPTER 39 EUV

苹果并不是半导体行业中唯一一家拥有复杂得令人眼花缭乱的供应链的公司到 2010 年代后期,荷兰光刻公司 ASML 已经花费了近 20 年的时间来尝试制作极紫外光刻技术。这样做需要在世界各地搜寻最先进的组件、最纯净的金属、最强大的激光和最精确的传感器。EUV 是我们这个时代最大的技术赌博之一。2012 年,在 ASML 生产出功能性 EUV 工具的几年前,英特尔、三星和台积电各自直接投资于 ASML,以确保该公司拥有继续开发其未来芯片制造能力所需的 EUV 工具所需的资金。仅英特尔2012 年向 ASML 投资了 40 亿美元,这是该公司有史以来最大的赌注之一,此前英特尔在 EUV 上投入了数十亿美元的赠款和投资,可追溯到 Andy Grove 时代。

Apple isn’t the only company in the semiconductor business with a bewilderingly complex supply chain. By the late-2010s, ASML, the Dutch lithography company, had spent nearly two decades trying to make extreme-ultraviolet lithography work. Doing so required scouring the world for the most advanced components, the purest metals, the most powerful lasers, and the most precise sensors. EUV was one of the biggest technological gambles of our time. In 2012, years before ASML had produced a functional EUV tool, Intel, Samsung, and TSMC had each invested directly in ASML to ensure the company had the funding needed to continue developing EUV tools that their future chipmaking capabilities would require. Intel alone invested $4 billion in ASML in 2012, one of the highest-stakes bets the company ever made, an investment that followed billions of dollars of previous grants and investments Intel had spent on EUV, dating back to the era of Andy Grove.

EUV 光刻工具背后的想法与英特尔和其他芯片公司财团为美国的几个国家实验室提供“感觉像是无限的资金解决一个不可能的问题,”正如参与该项目的一位科学家所说。这个概念与 Jay Lathrop 的倒置显微镜大致相同:通过使用“面具”创建光波图案阻挡一些光线,然后将光线投射到涂在硅片上的光刻胶化学物质上。光与光刻胶发生反应,可以沉积材料或将其蚀刻成完美的形状,从而产生工作芯片。

The idea behind EUV lithography tools was little changed from when Intel and a consortium of other chip firms had given several of America’s national labs “what felt like infinite money for solving an impossible problem,” as one of the scientists who worked on the project put it. The concept remained much the same as Jay Lathrop’s upside-down microscope: create a pattern of light waves by using a “mask” to block some of the light, then project the light onto photoresist chemicals applied to a silicon wafer. The light reacts with photoresists, making it possible to deposit material or etch it away in perfectly formed shapes, producing a working chip.

Lathrop 使用了简单的可见光和柯达生产的现成光刻胶。使用更复杂的透镜和化学品,最终可以在硅晶片上打印小至几百纳米的形状。可见光的波长本身是几百纳米,具体取决于颜色,因此随着晶体管变得越来越小,它最终面临着限制。该行业后来转向波长为 248 和 193 纳米的不同类型的紫外线。这些波长可以雕刻出比可见光更精确的形状,但它们也有局限性,因此业界将希望寄托在波长为 13.5 纳米的极紫外光上。

Lathrop had used simple visible light and off-the-shelf photoresists produced by Kodak. Using more complex lenses and chemicals, it eventually became possible to print shapes as small as a couple hundred nanometers on silicon wafers. The wavelength of visible light is itself several hundred nanometers, depending on the color, so it eventually faced limits as transistors were made ever smaller. The industry later moved to different types of ultraviolet light with wavelengths of 248 and 193 nanometers. These wavelengths could carve shapes more precise than visible light, but they, too, had limits, so the industry placed its hope on extreme ultraviolet light with a wavelength of 13.5 nanometers.

使用 EUV 光引入了几乎不可能解决的新困难。在 Lathrop 使用柯达生产的显微镜、可见光和光刻胶的地方,所有关键的 EUV 组件都必须专门制造。您不能简单地购买 EUV 灯泡。产生足够的 EUV 光需要用激光粉碎一个小锡球。Cymer 是一家由加州大学圣地亚哥分校的两位激光专家创立的公司,自 1980 年代以来一直是光刻光源的主要参与者。该公司的工程师意识到最好的方法是发射一个微小的锡球,该锡球以大约每小时 200 英里的速度在真空中移动,宽度为百万分之三米。然后用激光击打锡两次,第一个脉冲加热它,第二次将其爆破成温度约为 50 万度的等离子体,比太阳表面的温度高许多倍。然后,这种喷锡过程每秒重复五万次,以产生制造芯片所需数量的 EUV 光。Jay Lathrop 的光刻工艺依赖于一个简单的灯泡作为光源。从那时起,复杂性的增加令人难以置信。

Using EUV light introduced new difficulties that proved almost impossible to resolve. Where Lathrop used a microscope, visible light, and photoresists produced by Kodak, all the key EUV components had to be specially created. You can’t simply buy an EUV lightbulb. Producing enough EUV light requires pulverizing a small ball of tin with a laser. Cymer, a company founded by two laser experts from the University of California, San Diego, had been a major player in lithographic light sources since the 1980s. The company’s engineers realized the best approach was to shoot a tiny ball of tin measuring thirty-millionths of a meter wide moving through a vacuum at a speed of around two hundred miles per hour. The tin is then struck twice with a laser, the first pulse to warm it up, the second to blast it into a plasma with a temperature around half a million degrees, many times hotter than the surface of the sun. This process of blasting tin is then repeated fifty thousand times per second to produce EUV light in the quantities necessary to fabricate chips. Jay Lathrop’s lithography process had relied on a simple bulb for a light source. The increase in complexity since then was mind-boggling.

不过,Cymer 的光源只能工作,这要归功于一种新的激光,它可以以足够的功率粉碎锡滴。这需要一种比以前存在的任何一种更强大的基于二氧化碳的激光。2005 年夏天,Cymer 的两名工程师联系了一家名为 Trumpf 的德国精密工具公司,看看它是否可以制造这样的激光器。Trumpf 已经为精密切割等工业用途制造了世界上最好的二氧化碳激光器。这些激光器是德国最好的工业传统中加工的丰碑。因为二氧化碳激光器产生的大约 80% 的能量是热量,而只有 20% 是光,所以从机器中提取热量是一项关键挑战。Trumpf 之前设计了一个带有风扇的鼓风机系统,该系统每秒转动一千次,速度太快,无法依靠物理轴承。反而,将热量从激光系统中吸出,而不会磨损其他组件并危及可靠性。

Cymer’s light source only worked, though, thanks to a new laser that could pulverize the tin droplets with sufficient power. This required a carbon dioxide−based laser more powerful than any that previously existed. In summer 2005, two engineers at Cymer approached a German precision tooling company called Trumpf to see if it could build such a laser. Trumpf already made the world’s best carbon dioxide−based lasers for industrial uses like precision cutting. These lasers were monuments of machining in the best German industrial tradition. Because around 80 percent of the energy a carbon dioxide laser produces is heat and only 20 percent light, extracting heat from the machine is a key challenge. Trumpf had previously devised a system of blowers with fans that turned a thousand times a second, too fast to rely on physical bearings. Instead, the company learned to use magnets, so the fans floated in air, sucking heat out of the laser system without grinding against other components and imperiling reliability.

Trumpf 在提供 Cymer 所需的精度和可靠性方面享有盛誉和良好的记录。它可以提供动力吗?用于 EUV 的激光器需要比通快已经生产的激光器更强大。此外,Cymer 要求的精度比 Trumpf 之前处理的任何东西都要严格。该公司提出了一种由四个组件组成的激光器:两个“种子”激光器,功率低,但每个脉冲的时间准确,这样激光器每秒可以击中 5000 万个锡滴;四个增加光束功率的谐振器;一个超精确的“光束传输系统”,可将超过 30 米的光束导向锡滴室;以及确保激光直接命中的最终聚焦装置,每秒数百万次。

Trumpf had a reputation and a track record for delivering the precision and reliability Cymer needed. Could it deliver the power? Lasers for EUV needed to be substantially more powerful than the lasers Trumpf already produced. Moreover, the precision Cymer demanded was more exacting than anything Trumpf had previously dealt with. The company proposed a laser with four components: two “seed” lasers that are low power but accurately time each pulse so that the laser can hit 50 million tin drops a second; four resonators that increase the beam’s power; an ultra-accurate “beam transport system” that directs the beam over thirty meters toward the tin droplet chamber; and a final focusing device to ensure the laser scores a direct hit, millions of times a second.

每一步都需要新的创新。激光室中的特殊气体必须保持恒定的密度。锡滴本身会反射光,这可能会反射回激光器并干扰系统;为了防止这种情况,需要特殊的光学器件。该公司需要工业钻石来提供激光离开腔室的“窗口”,并且必须与合作伙伴开发新的超纯钻石。Trumpf 花了十年时间来应对这些挑战并生产出具有足够功率和可靠性的激光器。每一个都需要457,329 个零部件。

Every step required new innovations. Specialized gases in the laser chamber had to be kept at constant densities. The tin droplets themselves reflected light, which threatened to shine back into the laser and interfere with the system; to prevent this, special optics were required. The company needed industrial diamonds to provide the “windows” through which the laser exited the chamber, and had to work with partners to develop new, ultra-pure diamonds. It took Trumpf a decade to master these challenges and produce lasers with sufficient power and reliability. Each one required exactly 457,329 component parts.

在 Cymer 和 Trumpf 找到一种爆破锡的方法后,它可以发出足够的 EUV 光,下一步是制造镜子来收集光并将其引导到硅芯片上。蔡司是一家制造世界上最先进光学系统的德国公司,自 Perkin Elmer 和 GCA 时代以来,它就为光刻系统制造了反射镜和透镜。然而,过去使用的光学器件与 EUV 所需的光学器件之间的差异与 Lathrop 的灯泡和 Cymer 的喷锡液滴系统之间的对比一样巨大。

After Cymer and Trumpf found a way to blast tin so it emits sufficient EUV light, the next step was to create mirrors that collected the light and directed it toward a silicon chip. Zeiss, the German company that builds the world’s most advanced optical systems, had built mirrors and lenses for lithography systems since the days of Perkin Elmer and GCA. The difference between the optics used in the past and those required by EUV, however, was about as vast as the contrast between Lathrop’s lightbulb and Cymer’s system of blasting tin droplets.

Zeiss 的主要挑战是 EUV 难以反映。EUV 的 13.5nm 波长更接近 X 射线而不是可见光,并且与 X 射线一样,许多材料吸收 EUV 而不是反射它。蔡司开始开发由一百个钼和硅交替层制成的镜子,每层都有几纳米厚。劳伦斯利弗莫尔国家实验室的研究人员在 1998 年发表的一篇论文中将其确定为最佳 EUV 反射镜,但事实证明几乎不可能制造出这种具有纳米级精度的反射镜。最终,蔡司创造出的镜子是有史以来最光滑的物体,杂质几乎难以察觉。该公司表示,如果将 EUV 系统中的镜子缩放到德国的大小,它们最大的不规则性将是十分之一毫米。为了精确引导 EUV 光,它们必须保持完全静止,需要如此精确的机械装置和传感器,以至于蔡司吹嘘它们可以用来将激光瞄准打一个远到月球的高尔夫球。

Zeiss’s primary challenge was that EUV is difficult to reflect. The 13.5nm wavelength of EUV is closer to X-rays than to visible light, and as is the case with X-rays, many materials absorb EUV rather than reflect it. Zeiss began developing mirrors made of one hundred alternating layers of molybdenum and silicon, each layer a couple nanometers thick. Researchers in Lawrence Livermore National Lab had identified this as an optimal EUV mirror in a paper published in 1998, but building such a mirror with nanoscale precision proved almost impossible. Ultimately, Zeiss created mirrors that were the smoothest objects ever made, with impurities that were almost imperceptibly small. If the mirrors in an EUV system were scaled to the size of Germany, the company said, their biggest irregularities would be a tenth of a millimeter. To direct EUV light with precision, they must be held perfectly still, requiring mechanics and sensors so exact that Zeiss boasted they could be used to aim a laser to hit a golf ball as far away as the moon.

对于 2013 年接管 ASML EUV 业务的 Frits van Houts 来说,对 EUV 光刻系统最关键的投入不是任何单个组件,而是公司自身在供应链管理方面的技能。ASML 设计了这种业务关系网络“就像一台机器,”van Houts 解释说,生产了一个由数千家公司组成的微调系统,能够满足 ASML 的严格要求。ASML 本身只生产了 15%他估计,一个 EUV 工具的组件的其余部分是从其他公司购买的。这让它能够接触到世界上最精良的产品,但它也需要不断的监视。

For Frits van Houts, who took over leadership of ASML’s EUV business in 2013, the most crucial input into an EUV lithography system wasn’t any individual component, but the company’s own skill in supply chain management. ASML engineered this network of business relationships “like a machine,” van Houts explained, producing a finely tuned system of several thousand companies capable of meeting ASML’s exacting requirements. ASML itself only produced 15 percent of an EUV tool’s components, he estimated, buying the rest from other firms. This let it access the world’s most finely engineered goods, but it also required constant surveillance.

该公司别无选择,只能依靠单一来源获取 EUV 系统的关键组件。为了解决这个问题,ASML 深入了解其供应商的供应商以了解风险。ASML 通过投资奖励某些供应商,例如它在 2016 年向蔡司支付了 10 亿美元,用于资助该公司的研发过程。然而,它使所有这些都达到了严格的标准。“如果你不守规矩,我们就会收买你,”ASML 的首席执行官 Peter Wennink 告诉一位供应商。这不是开玩笑:ASML 最终收购了包括 Cymer 在内的多家供应商,因为他们认为自己可以更好地管理它们。

The company had no choice but to rely on a single source for the key components of an EUV system. To manage this, ASML drilled down into its suppliers’ suppliers to understand the risks. ASML rewarded certain suppliers with investment, like the $1 billion it paid Zeiss in 2016 to fund that company’s R&D process. It held all of them, however, to exacting standards. “If you don’t behave, we’re going to buy you,” ASML’s CEO Peter Wennink told one supplier. It wasn’t a joke: ASML ended up buying several suppliers, including Cymer, after concluding it could better manage them itself.

结果是一台拥有数十万个组件的机器,花费了数百亿美元和数十年的时间来开发。奇迹不仅仅在于 EUV 光刻技术有效,还在于它的可靠性足以以经济高效的方式生产芯片。极高的可靠性对于将被放入 EUV 系统的任何组件都至关重要。ASML 为每个组件设定了一个目标,即平均至少持续三万小时——大约四年——在需要维修之前。在实践中,需要更频繁地进行维修,因为并非每个零件都同时损坏。每台 EUV 机器的成本超过 1 亿美元,因此每离线一小时,芯片制造商就会损失数千美元的生产成本。

The result was a machine with hundreds of thousands of components that took tens of billions of dollars and several decades to develop. The miracle isn’t simply that EUV lithography works, but that it does so reliably enough to produce chips cost-effectively. Extreme reliability was crucial for any component that would be put in the EUV system. ASML had set a target for each component to last on average for at least thirty thousand hours—around four years—before needing repair. In practice, repairs would be needed more often, because not every part breaks at the same time. EUV machines cost over $100 million each, so every hour one is offline costs chipmakers thousands of dollars in lost production.

EUV 工具之所以有效,部分原因是它们的软件有效。例如,ASML 使用预测性维护算法来猜测组件何时需要在损坏之前更换。它还使用用于称为计算光刻的过程的软件来更准确地打印图案。光波与光刻胶化学物质反应的原子级不可预测性给 EUV 带来了新问题,而大波长光刻几乎不存在这些问题。为了调整光线折射方式的异常,ASML 的工具以不同于芯片制造商希望印在芯片上的图案投射光线。打印“X”需要使用形状非常不同的图案,但最终会在光波撞击硅晶片时产生“X”。

EUV tools work in part because their software works. ASML uses predictive maintenance algorithms to guess when components need to be replaced before they break, for example. It also uses software for a process called computational lithography to print patterns more exactly. The atomic-level unpredictability in light waves’ reaction with photoresist chemicals created new problems with EUV that barely existed with larger-wavelength lithography. To adjust for anomalies in the way light refracts, ASML’s tools project light in a pattern that differs from what chipmakers want imprinted on a chip. Printing an “X” requires using a pattern with a very different shape but which ends up creating an “X” when the light waves hit the silicon wafer.

最终产品——芯片——之所以能如此可靠地工作,是因为它们只有一个组件:一块顶部覆盖有其他金属的硅块。芯片中没有移动部件,除非你数一下内部快速移动的电子。然而,生产先进的半导体依赖于一些有史以来最复杂的机器。ASML 的 EUV 光刻工具是历史上最昂贵的量产机床,如此复杂,如果没有 ASML 人员的广泛培训,就无法使用,他们在工具的整个生命周期内都留在现场。每个 EUV 扫描仪的侧面都有一个 ASML 标志。但该公司欣然承认,ASML 的专长在于其能够协调由光学专家、软件设计师、激光公司和许多其他人组成的广泛网络,这些人的能力是实现 EUV 梦想所需的。

The final product—chips—work so reliably because they only have a single component: a block of silicon topped with other metals. There are no moving parts in a chip, unless you count the electrons zipping around inside. Producing advanced semiconductors, however, has relied on some of the most complex machinery ever made. ASML’s EUV lithography tool is the most expensive mass-produced machine tool in history, so complex it’s impossible to use without extensive training from ASML personnel, who remain on-site for the tool’s entire life span. Each EUV scanner has an ASML logo on its side. But ASML’s expertise, the company readily admits, was its ability to orchestrate a far-flung network of optics experts, software designers, laser companies, and many others whose capabilities were needed to make the dream of EUV a reality.

就像安迪·格鲁夫在他生命的最后几年所做的那样,很容易为制造业的离岸外包感到遗憾。荷兰公司 ASML 将一项由美国国家实验室开创并主要由英特尔资助的技术商业化,这无疑会激怒美国的经济民族主义者,如果他们知道光刻或 EUV 技术的历史的话。然而,ASML 的 EUV 工具并不是真正的荷兰人,尽管它们主要是在荷兰组装的。关键部件来自加利福尼亚的 Cymer 以及德国的 Zeiss 和 Trumpf。甚至这些德国公司也依赖美国生产的关键设备。关键是,不是一个国家能够对这些神奇的工具感到自豪,它们是许多国家的产物。具有数十万个零件的工具有许多父亲。

It’s easy to lament the offshoring of manufacturing, as Andy Grove did during the final years of his life. That a Dutch company, ASML, had commercialized a technology pioneered in America’s National Labs and largely funded by Intel would undoubtedly have rankled America’s economic nationalists, had any been aware of the history of lithography or of EUV technology. Yet ASML’s EUV tools weren’t really Dutch, though they were largely assembled in the Netherlands. Crucial components came from Cymer in California and Zeiss and Trumpf in Germany. And even these German firms relied on critical pieces of U.S.-produced equipment. The point is that, rather than a single country being able to claim pride of ownership regarding these miraculous tools, they are the product of many countries. A tool with hundreds of thousands of parts has many fathers.

“行得通吗?” Andy Grove 在将他的第一笔 2 亿美元投资于 EUV 之前,曾询问过 John Carruthers。经过三个十年的投资、数十亿美元、一系列的技术创新,以及世界上最复杂的供应链之一的建立,到 2010 年代中期,ASML 的 EUV 工具终于准备好部署在世界上最先进的芯片上晶圆厂。

“Will it work?” Andy Grove had asked John Carruthers, before investing his first $200 million in EUV. After three decades of investment, billions of dollars, a series of technological innovations, and the establishment of one of the world’s most complex supply chains, by the mid-2010s, ASML’s EUV tools were finally ready to be deployed in the world’s most advanced chip fabs.

第 40 章 “没有B计划”

CHAPTER 40 “There Is No Plan B”

2015 年,Tony Yen 被问到如果 ASML 正在开发的新型极紫外光刻工具不起作用会怎样。在过去的 25 年里,Yen 一直在光刻技术的前沿工作。1991 年,他刚从麻省理工学院被德州仪器公司聘用,在那里他修补了 GCA 在破产前生产的最终光刻工具之一。然后,他在 1990 年代后期加入台积电,当时正值生产波长为 193 纳米的光的深紫外光刻工具上线。近二十年来,该行业依靠这些工具制造更小的晶体管,使用一系列光学技巧,例如通过水或多个掩模发射光,以使测量 193nm 的光波能够图案化尺寸的一小部分。这些技巧使摩尔定律保持活力,

In 2015, Tony Yen was asked what would happen if the new extreme-ultraviolet lithography tool that ASML was developing didn’t work. Yen had spent the prior twenty-five years working at the cutting edge of lithography. In 1991 he’d been hired fresh out of MIT by Texas Instruments, where he tinkered with one of the final lithography tools GCA produced before going bankrupt. He then joined TSMC in the late 1990s just as deep-ultraviolet lithography tools, which produced light with a wavelength of 193 nanometers, were coming online. For nearly two decades, the industry relied on these tools to fabricate ever-smaller transistors, using a series of optical tricks like shooting light through water or through multiple masks to enable light waves measuring 193nm to pattern shapes a fraction of the size. These tricks kept Moore’s Law alive, as the chip industry shrank transistors from the 180nm node in the late 1990s through the early stages of 3D FinFET chips, which were ready for high-volume manufacturing by the mid-2010s.

然而,只有这么多的光学技巧可以帮助 193nm 光雕刻更小的特征。每个新的解决方法都会增加时间和成本。到 2010 年代中期,可能已经可以勉强进行一些额外的改进,但摩尔定律需要更好的光刻工具来雕刻更小的形状。唯一的希望是自 1990 年代初以来一直在开发中的极大延迟的 EUV 光刻工具终于可以在商业规模上使用。什么是替代方案?“没有 B 计划,”Yen 知道。

However, there were only so many optical tricks that could help 193nm light carve smaller features. Each new workaround added time and cost money. By the mid-2010s, it might have been possible to eke out a couple additional improvements, but Moore’s Law needed better lithography tools to carve smaller shapes. The only hope was that the hugely delayed EUV lithography tools, which had been in development since the early 1990s, could finally be made to work at a commercial scale. What was the alternative? “There is no Plan B,” Yen knew.

Morris Chang 在 EUV 上的赌注比半导体行业的任何人都多。该公司的光刻团队对 EUV 工具是否已准备好进行大批量制造存在分歧,但领导台积电研发并因​​公司一流的制造技术而广受赞誉的说话温和的工程师蒋尚义坚信 EUV 是前进的唯一道路。蒋出生在重庆,和张忠谋一样,他的家人在二战期间逃离了日本军队。他在台湾长大,之后在斯坦福大学学习电气工程,并在德克萨斯州的 TI 和硅谷的惠普找到工作。1997 年,当台积电突然打电话给他工作机会和巨额签约奖金时,他搬回台湾帮助建立公司。2006年,他尝试在加利福尼亚退休,

Morris Chang bet more heavily on EUV than anyone else in the semiconductor industry. The company’s lithography team was divided over whether EUV tools were ready for high-volume manufacturing, but Shang-yi Chiang, the soft-spoken engineer who headed TSMC’s R&D and was widely credited for the company’s top-notch manufacturing technology, was convinced EUV was the only path forward. Chiang was born in Chongqing where, like Morris Chang, his family had fled from Japanese armies during World War II. He grew up in Taiwan before studying electrical engineering at Stanford and landing jobs at TI in Texas and then at HP in Silicon Valley. When TSMC called out of the blue with a job offer—and a massive signing bonus—he moved back to Taiwan in 1997 to help build the company. In 2006, he tried retiring in California, but when TSMC faced a delay with its 40nm manufacturing process in 2009, a frustrated Morris Chang ordered Chiang back to Taiwan and over a meal of beef noodle soup asked him to again take up the responsibility of managing R&D.

曾在德克萨斯州和加利福尼亚州以及台湾工作过的蒋总是对推动台积电的雄心和职业道德感到震惊。这一雄心源于张忠谋对世界领先技术的愿景,他愿意斥巨资将台积电的研发团队从 1997 年的 120 人扩大到 2013 年的 7000 人。这种渴望渗透到整个公司。“在台湾,人们工作得更加努力,”蒋解释道。由于制造工具占先进晶圆厂的大部分成本,因此保持设备运行对于盈利能力至关重要。蒋说,在美国,如果凌晨 1 点出了问题,工程师第二天早上就会修好。在台积电,他们会在凌晨 2 点之前修复它“他们不会抱怨,”他解释说,并且“他们的配偶也没有抱怨”。随着蒋重新负责研发,台积电向EUV迈进。他毫不费力地找到员工通宵工作。他要求在公司最大的设施之一 Fab 12 的中间建造了三台用于测试目的的 EUV 扫描仪,并且在公司与 ASML 的合作中,他不遗余力地测试和改进 EUV 工具。

Having worked in Texas and California as well as in Taiwan, Chiang was always struck by the ambition and the work ethic that drove TSMC. The ambition stemmed from Morris Chang’s vision of world-beating technology, evident in his willingness to spend huge sums expanding TSMC’s R&D team from 120 people in 1997 to 7,000 in 2013. This hunger permeated the entire company. “People worked so much harder in Taiwan,” Chiang explained. Because manufacturing tools account for much of the cost of an advanced fab, keeping the equipment operating is crucial for profitability. In the U.S., Chiang said, if something broke at 1 a.m., the engineer would fix it the next morning. At TSMC, they’d fix it by 2 a.m. “They do not complain,” he explained, and “their spouse does not complain” either. With Chiang back in charge of R&D, TSMC charged forward toward EUV. He had no difficulty finding employees to work all night long. He requested that three EUV scanners for testing purposes be built in the middle of one of the company’s biggest facilities, Fab 12, and in the company’s partnership with ASML he spared no expense in testing and improving EUV tools.

与台积电、三星和英特尔一样,GlobalFoundries 正在考虑采用 EUV,因为它为自己的 7nm 节点做准备。从创立之初,GlobalFoundries 就知道,如果要发展壮大,就需要发展壮大。该公司继承了 AMD 的晶圆厂,但规模远小于其竞争对手。为了成长,GlobalFoundries 拥有2010 年收购了总部位于新加坡的代工厂特许半导体公司。几年后,也就是 2014 年,它收购了 IBM 的微电子业务,承诺为蓝色巨人生产芯片,后者出于与 AMD 相同的原因决定无晶圆厂。IBM 高管曾经分享过计算生态系统的形象:一个倒置的金字塔,底部有半导体,所有其他计算都依赖于它。然而,尽管 IBM 在半导体业务的增长中发挥了重要作用,但其领导人得出结论认为,制造芯片没有经济意义。面对投资数十亿建设新的先进晶圆厂或数十亿高利润软件的决定,他们选择了后者,将其芯片部门出售给 GlobalFoundries。

Like TSMC, Samsung, and Intel, GlobalFoundries was considering adopting EUV as it prepared for its own 7nm node. From its creation, GlobalFoundries knew it needed to grow if it was to thrive. The company had inherited AMD’s fabs, but it was far smaller than its rivals. To grow, GlobalFoundries had bought Chartered Semiconductor, a Singapore-based foundry, in 2010. Several years later, in 2014, it bought IBM’s microelectronics business, promising to produce chips for Big Blue, which had decided to go fabless for the same reason as AMD. IBM executives used to share an image of the computing ecosystem: an upside-down pyramid with semiconductors at the bottom, on which all other computing depended. Yet though IBM had played a fundamental role in the growth of the semiconductor business, its leaders concluded that fabricating chips made no financial sense. Facing a decision to invest billions to build a new advanced fab, or billions on high-margin software, they chose the latter, selling their chip division to GlobalFoundries.

到 2015 年,得益于这些收购,GlobalFoundries 是迄今为止美国最大的代工厂,也是世界上最大的代工厂之一,但与台积电相比,它仍然是一个小鱼。GlobalFoundries 与台湾联华电子争夺全球第二大晶圆代工厂的地位,每家公司拥有约10% 的代工市场。然而,台积电拥有全球 50% 以上的代工市场。三星在 2015 年仅占有 5% 的晶圆代工市场,但将其内部设计的大量芯片(例如,用于智能手机处理器的内存芯片和芯片)的大量生产包括在内,它生产的晶圆比任何人都多。按照行业标准每月数千片晶圆来衡量,台积电的产能为 180 万片,而三星则为 250 万片。格罗方德只有700,000。

By 2015, thanks to these acquisitions, GlobalFoundries was by far the biggest foundry in the United States and one of the largest in the world, but it was still a minnow compared to TSMC. GlobalFoundries competed with Taiwan’s UMC for status as the world’s second-largest foundry, with each company having about 10 percent of the foundry marketplace. However, TSMC had over 50 percent of the world’s foundry market. Samsung only had 5 percent of the foundry market in 2015, but it produced more wafers than anyone when its vast production of chips designed in-house (for example, memory chips and chips for smartphone processors) were included. Measured by thousands of wafers per month, the industry standard, TSMC had a capacity of 1.8 million while Samsung had 2.5 million. GlobalFoundries had only 700,000.

台积电、英特尔和三星肯定会采用 EUV,尽管它们在何时以及如何采用方面有不同的策略。格罗方德没有那么自信。该公司一直在努力应对其 28nm 工艺。为了减少延误的风险,它决定从三星授权其 14nm 工艺,而不是内部开发,这一决定并不表明对其研发工作的信心。

TSMC, Intel, and Samsung were certain to adopt EUV, though they had different strategies about when and how to embrace it. GlobalFoundries was less confident. The company had struggled with its 28nm process. To reduce the risk of delays, it decided to license its 14nm process from Samsung rather than develop it in-house, a decision that didn’t suggest confidence in its R&D efforts.

到 2018 年,GlobalFoundries 已经购买了几台 EUV 光刻工具,并在其最先进的设施 Fab 8 中安装它们,但当时公司高管下令停止工作。EUV 计划被取消。GlobalFoundries 正在放弃生产新的尖端节点。它不会采用基于 EUV 光刻技术的 7nm 工艺,该工艺已经花费了 15 亿美元的开发成本,并且需要相当数量的额外支出才能上线。台积电、英特尔和三星的财务状况足够强大,可以掷骰子,并希望他们能够让 EUV 发挥作用。GlobalFoundries 认为,作为一家中型代工厂,它永远无法让 7nm 工艺在财务上可行。它宣布将停止制造更小的晶体管,将研发支出削减三分之一,并在多年亏损后迅速扭亏为盈。除了世界上最大的芯片制造商之外,制造尖端处理器对所有人来说都太昂贵了。即使是拥有 GlobalFoundries 的波斯湾皇室成员的财力雄厚,也不够雄厚。能够制造尖端逻辑芯片的公司数量从四家下降到三家。

By 2018, GlobalFoundries had purchased several EUV lithography tools and was installing them in its most advanced facility, Fab 8, when the company’s executives ordered them to halt work. The EUV program was being canceled. GlobalFoundries was giving up production of new, cutting-edge nodes. It wouldn’t pursue a 7nm process based on EUV lithography, which had already cost $1.5 billion in development and would have required a comparable amount of additional spending to bring online. TSMC, Intel, and Samsung had financial positions that were strong enough to roll the dice and hope they could make EUV work. GlobalFoundries decided that as a medium-sized foundry, it could never make a 7nm process financially viable. It announced it would stop building ever-smaller transistors, slashed R&D spending by a third, and quickly turned a profit after several years of losses. Building cutting-edge processors was too expensive for everyone except the world’s biggest chipmakers. Even the deep pockets of the Persian Gulf royals who owned GlobalFoundries weren’t deep enough. The number of companies capable of fabricating leading-edge logic chips fell from four to three.

第 41 章 英特尔如何忘记创新

CHAPTER 41 How Intel Forgot Innovation

一个至少美国可以指望英特尔。该公司在半导体行业拥有无与伦比的地位。旧的领导层早已不复存在——安迪·格罗夫于 2016 年去世,而戈登·摩尔,现年 90 岁,退休到夏威夷——但将 DRAM 商业化并发明了微处理器的名声依然存在。没有哪家公司在将创新芯片设计与制造能力相结合时有着更好的业绩记录。英特尔的 x86 架构仍然是 PC 和数据中心的行业标准。个人电脑市场停滞不前,因为似乎几乎每个人都已经拥有了个人电脑,但它对英特尔来说仍然非常有利可图,每年提供数十亿美元可以再投资于研发。在整个 2010 年代,该公司每年在研发上花费超过 100 亿美元,是台积电的四倍,是 DARPA 整个预算的三倍。世界上只有几家公司花费更多。

At least the United States could count on Intel. The company had an unparalleled position in the semiconductor industry. The old leadership was long gone—Andy Grove died in 2016, while Gordon Moore, now in his nineties, retired to Hawaii—but the reputation of having commercialized the DRAM and invented the microprocessor remained. No company had a better track record combining innovative chip design with manufacturing prowess. Intel’s x86 architecture remained the industry standard for PCs and data centers. The PC market was stagnant, because it seemed nearly everyone already had a PC, but it remained remarkably profitable for Intel, providing billions of dollars a year that could be reinvested into R&D. The company spent over $10 billion a year on R&D throughout the 2010s, four times as much as TSMC and three times more than the entire budget of DARPA. Only a couple of companies in the world spent more.

随着芯片行业进入 EUV 时代,英特尔似乎蓄势待发。这家公司对 EUV 的出现至关重要,这要归功于安迪·格鲁夫在 1990 年代初期对该技术的 2 亿美元投资。现在,经过数十亿美元的投资(其中很大一部分来自英特尔),ASML 终于使这项技术成为现实。然而,英特尔并没有利用这个缩小晶体管的新时代,而是浪费了它的领先优势,错过了主要的人工智能所需的半导体架构发生了变化,然后导致制造工艺混乱,无法跟上摩尔定律。

As the chip industry entered the EUV era, Intel looked poised to dominate. The company had been crucial to EUV’s emergence, thanks to Andy Grove’s initial $200 million bet on the technology in the early 1990s. Now, after billions of dollars of investment—a substantial portion of which had come from Intel—ASML had finally made the technology a reality. Yet rather than capitalizing on this new era of shrinking transistors, Intel squandered its lead, missing major shifts in semiconductor architecture needed for artificial intelligence, then bungling its manufacturing processes and failing to keep up with Moore’s Law.

英特尔今天仍然保持着巨大的利润。它仍然是美国最大、最先进的芯片制造商。然而,自从格鲁夫在 1980 年代决定放弃内存并将一切都押在微处理器上以来,它的未来比任何时候都更令人怀疑。它仍然有机会在未来五年内重新获得领导地位,但它也很容易最终倒闭。危在旦夕的不仅仅是一家公司,而是美国芯片制造行业的未来。没有英特尔,就没有一家美国公司——或者台湾或韩国以外的一家工厂——能够制造尖端处理器。

Intel remains enormously profitable today. It’s still America’s biggest and most advanced chipmaker. However, its future is more in doubt than at any point since Grove’s decision in the 1980s to abandon memory and bet everything on microprocessors. It still has a shot at regaining its leadership position over the next half decade, but it could just as easily end up defunct. What’s at stake isn’t simply one company, but the future of America’s chip fabrication industry. Without Intel, there won’t be a single U.S. company—or a single facility outside of Taiwan or South Korea—capable of manufacturing cutting-edge processors.

英特尔在 2010 年代成为硅谷的一个异类。大多数美国最大的逻辑芯片市场公司,包括英特尔的主要竞争对手 AMD,都卖掉了他们的晶圆厂,只专注于设计。英特尔顽固地坚持其集成模式——将半导体设计和制造结合在一家公司中——那里的高管认为这仍然是生产芯片的最佳方式。英特尔的领导人认为,公司的设计和制造流程相互优化。相比之下,台积电别无选择,只能采用通用制造工艺,这种工艺对于高通智能手机处理器和 AMD 服务器芯片同样适用。

Intel entered the 2010s as an outlier in Silicon Valley. Most of America’s biggest firms in the market for logic chips, including Intel’s archrival AMD, had sold their fabs and focused only on design. Intel stuck stubbornly to its integrated model—combining semiconductor design and manufacturing in one company—which executives there thought was still the best way to churn out chips. The company’s design and manufacturing processes were optimized for each other, Intel’s leaders argued. TSMC, by contrast, had no choice but to adopt generic manufacturing processes that could work just as well for a Qualcomm smartphone processor as an AMD server chip.

英特尔认为集成模型的一些好处是正确的,但也有很大的缺点。由于台积电为许多不同的公司制造芯片,它现在每年制造的硅片数量几乎是英特尔的三倍,因此它拥有有更多机会磨练它的过程。此外,英特尔将芯片设计初创公司视为威胁,台积电则看到了制造服务的潜在客户。因为台积电只有一个价值主张——有效制造——它的领导地位一直专注于以更低的成本制造更先进的半导体。英特尔的领导人不得不将注意力分散在芯片设计和芯片制造之间。他们最终都搞砸了。

Intel was right to perceive some benefits of an integrated model, but there were substantial downsides. Because TSMC manufactures chips for many different companies, it now fabricates nearly three times as many silicon wafers per year as Intel, so it has more chance to hone its process. Moreover, where Intel saw chip design startups as a threat, TSMC saw potential customers for manufacturing services. Because TSMC had only a single value proposition—effective manufacturing—its leadership focused relentlessly on fabricating ever-more-advanced semiconductors at lower cost. Intel’s leaders had to split their attention between chip design and chip manufacturing. They ended up bungling both.

英特尔的第一个问题是人工智能。到 2010 年代初,该公司的核心市场,即供应 PC 处理器,已经停滞不前。如今,除了游戏玩家之外,几乎没有人会在新型号发布时兴奋地升级他们的 PC,而且大多数人也不会过多考虑内部的处理器类型。英特尔的另一个主要市场——为数据中心的服务器销售处理器——在 2010 年代蓬勃发展。亚马逊网络服务、微软 Azure、谷歌云和其他公司构建了庞大的数据中心网络,这些网络提供了使“云”成为可能的计算能力。我们在线使用的大部分数据都是在这些公司的数据中心之一处理的,每个数据中心都装满了英特尔芯片。但在 2010 年代初期,就在英特尔完成对数据中心的征服之际,处理需求开始发生转变。

Intel’s first problem was artificial intelligence. By the early 2010s, the company’s core market, supplying PC processors, had stalled. Today, other than gamers, hardly anyone excitedly upgrades their PC when a new model is released, and most people don’t think much about which type of processor is inside. Intel’s other main market—selling processors for servers in data centers—boomed over the 2010s. Amazon Web Services, Microsoft Azure, Google Cloud, and other companies built networks of vast data centers, which provided the computing power that made possible “the cloud.” Most of the data we use online is processed in one of these companies’ data centers, each of which is full of Intel chips. But in the early 2010s, just as Intel completed its conquest of the data center, processing demands began to shift. The new trend was artificial intelligence—a task that Intel’s main chips were poorly designed to address.

自 1980 年代以来,英特尔专注于一种称为 CPU 的芯片,即中央处理单元,PC 中的微处理器就是其中的一个例子。这些是在计算机或数据中心中充当“大脑”的芯片。它们是通用的主力,同样能够打开 Web 浏览器或运行 Microsoft Excel。它们可以进行许多不同类型的计算,这使得它们具有多种用途,但它们是一个接一个地连续进行这些计算。

Since the 1980s, Intel has specialized in a type of chip called a CPU, a central processing unit, of which a microprocessor in a PC is one example. These are the chips that serve as the “brain” in a computer or data center. They are general-purpose workhorses, equally capable of opening a web browser or running Microsoft Excel. They can conduct many different types of calculations, which makes them versatile, but they do these calculations serially, one after another.

可以在通用 CPU 上运行任何 AI 算法,但 AI 所需的计算规模使得使用 CPU 的成本过高。训练单个 AI 模型的成本——它使用的芯片和它们消耗的电力——可以延伸到数百万美元。(要训练计算机识别猫,你必须向它展示很多猫和狗,以便它学会区分两者。你的算法需要的动物越多,你需要的晶体管就越多。)

It’s possible to run any AI algorithm on a general-purpose CPU, but the scale of computation required for AI makes using CPUs prohibitively expensive. The cost of training a single AI model—the chips it uses and the electricity they consume—can stretch into the millions of dollars. (To train a computer to recognize a cat, you have to show it a lot of cats and dogs so it learns to distinguish between the two. The more animals your algorithm requires, the more transistors you need.)

由于 AI 工作负载通常需要重复运行相同的计算,每次使用不同的数据,因此找到一种为 AI 算法定制芯片的方法对于使其在经济上可行至关重要。亚马逊和微软等大型云计算公司运营着大多数公司算法运行的数据中心,每年花费数百亿美元购买芯片和服务器。他们还花费巨资为这些数据中心供电。在他们竞相向公司出售其“云”中的空间时,从他们的芯片中榨取效率是必要的。与通用 Intel CPU 相比,针对 AI 优化的芯片可以更快地工作、占用更少的数据中心空间并使用更少的功率。

Because AI workloads often require running the same calculation repeatedly, using different data each time, finding a way to customize chips for AI algorithms is crucial to making them economically viable. Big cloud computing companies like Amazon and Microsoft, which operate the data centers on which most companies’ algorithms run, spend tens of billions of dollars annually buying chips and servers. They also spend vast sums providing electricity for these data centers. Wringing efficiencies out of their chips is a necessity as they compete to sell companies space in their “cloud.” Chips optimized for AI can work faster, take up less data center space, and use less power than general-purpose Intel CPUs.

在 2010 年代初,图形芯片的设计者 Nvidia 开始听到有关斯坦福大学博士生使用 Nvidia 的图形处理单元 (GPU) 来处理图形以外的事情的传闻。GPU 的设计与标准 Intel 或 AMD CPU 的工作方式不同,后者具有无限的灵活性,但会一个接一个地运行所有计算。相比之下,GPU 旨在一次运行同一计算的多次迭代。这种类型的“并行处理”很快就变得很清楚,它的用途不仅仅是控制计算机游戏中的图像像素。它还可以有效地训练人工智能系统。CPU 会一个接一个地为算法提供多条数据,而 GPU 可以同时处理多条数据。为了学习识别猫的图像,CPU 会逐个像素地处理,而 GPU 可以同时“查看”多个像素。因此,训练计算机识别猫所需的时间急剧减少。

In the early 2010s, Nvidia—the designer of graphic chips—began hearing rumors of PhD students at Stanford using Nvidia’s graphics processing units (GPUs) for something other than graphics. GPUs were designed to work differently from standard Intel or AMD CPUs, which are infinitely flexible but run all their calculations one after the other. GPUs, by contrast, are designed to run multiple iterations of the same calculation at once. This type of “parallel processing,” it soon became clear, had uses beyond controlling pixels of images in computer games. It could also train AI systems efficiently. Where a CPU would feed an algorithm many pieces of data, one after the other, a GPU could process multiple pieces of data simultaneously. To learn to recognize images of cats, a CPU would process pixel after pixel, while a GPU could “look” at many pixels at once. So the time needed to train a computer to recognize cats decreased dramatically.

此后,英伟达将其未来押在了人工智能上。从成立之初,英伟达就将其制造业务外包给了台积电,并坚持不懈地专注于设计新一代 GPU,并对其称为 CUDA 的特殊编程语言进行定期改进,这使得设计使用英伟达芯片的程序变得简单。随着投资者押注数据中心将需要更多的 GPU,英伟达已经成为美国最有价值的半导体公司。

Nvidia has since bet its future on artificial intelligence. From its founding, Nvidia outsourced its manufacturing, largely to TSMC, and focused relentlessly on designing new generations of GPUs and rolling out regular improvements to its special programming language called CUDA that makes it straightforward to devise programs that use Nvidia’s chips. As investors bet that data centers will require ever more GPUs, Nvidia has become America’s most valuable semiconductor company.

然而,它的崛起并不确定,因为除了购买 Nvidia 芯片之外,大型云公司——谷歌、亚马逊、微软、Facebook、腾讯、阿里巴巴等——也开始设计自己的芯片,专门满足他们的处理需求,专注于人工智能和机器学习。例如,谷歌有设计了自己的称为张量处理单元 (TPU) 的芯片,该芯片经过优化,可与 Google 的 TensorFlow 软件库配合使用。您可以在其爱荷华州数据中心以每月 3,000 美元的价格租用 Google 最简单的 TPU,但更强大的 TPU 的价格每月可以达到 100,000 美元以上。云可能听起来很空灵,但我们所有数据所在的硅片是非常真实的,而且非常昂贵。

Its ascent isn’t assured, however, because in addition to buying Nvidia chips the big cloud companies—Google, Amazon, Microsoft, Facebook, Tencent, Alibaba, and others—have also begun designing their own chips, specialized to their processing needs, with a focus on artificial intelligence and machine learning. For example, Google has designed its own chips called Tensor processing units (TPUs), which are optimized for use with Google’s TensorFlow software library. You can rent the use of Google’s simplest TPU in its Iowa data center for $3,000 per month, but prices for more powerful TPUs can reach over $100,000 monthly. The cloud may sound ethereal, but the silicon on which all our data lives is very real—and very expensive.

无论是英伟达还是大型云计算公司,英特尔在数据中心处理器销售方面的近乎垄断地位正在结束。如果英特尔找到了新市场,那么失去这一主导地位的问题就不会那么严重了。然而,该公司在 2010 年代中期进军代工业务,试图与台积电正面竞争,但以失败告终。英特尔试图向任何寻求芯片制造服务的客户开放其生产线,但他悄悄地承认,集成设计和制造的模式并不像英特尔高管所声称的那样成功。该公司具备成为主要代工企业的所有条件,包括先进的技术和庞大的生产能力,但要想成功,就需要进行重大的文化变革。台积电以知识产权开放,但英特尔是封闭和秘密的。台积电是以服务为导向的,而英特尔则认为客户应该遵循自己的规则。台积电没有与客户竞争,因为它没有设计任何芯片。英特尔是行业巨头,其芯片几乎与所有人竞争。

Whether it will be Nvidia or the big cloud companies doing the vanquishing, Intel’s near-monopoly in sales of processors for data centers is ending. Losing this dominant position would have been less problematic if Intel had found new markets. However, the company’s foray into the foundry business in the mid-2010s, where it tried to compete head-on with TSMC, was a flop. Intel tried opening its manufacturing lines to any customers looking for chipmaking services, quietly admitting that the model of integrated design and manufacturing wasn’t nearly as successful as Intel’s executives claimed. The company had all the ingredients to become a major foundry player, including advanced technology and massive production capacity, but succeeding would have required a major cultural change. TSMC was open with intellectual property, but Intel was closed off and secretive. TSMC was service-oriented, while Intel thought customers should follow its own rules. TSMC didn’t compete with its customers, since it didn’t design any chips. Intel was the industry giant whose chips competed with almost everyone.

2013 年至 2018 年担任英特尔 CEO 的布赖恩·科再奇公开坚称,“过去几年我基本上一直在经营我们的代工业务”,并将这项工作描述为“具有战略意义”。但在客户看来,情况并非如此,他们认为公司未能将代工客户放在首位。在英特尔内部,代工业务并未被视为优先事项。与制造 PC 和数据中心芯片(仍然是高利润业务)相比,新的代工企业拥有内部支持很少。因此,英特尔的代工业务在 2010 年代运营时只赢得了一个主要客户。它是仅仅几年后就关闭了。

Brian Krzanich, who was Intel’s CEO from 2013 to 2018, insisted publicly that “I’ve been basically running our foundry business for the last few years” and described the effort as “strategically important.” But it didn’t look that way to customers, who thought the company failed to put foundry customers first. Inside Intel, the foundry business wasn’t treated as a priority. Compared to making PC and data center chips—which remained highly profitable businesses—the new foundry venture had little internal support. So Intel’s foundry business won only a single major customer while in operation in the 2010s. It was shuttered after just several years.

随着英特尔在 2018 年接近其成立 50 周年,衰退已经开始。该公司的市场份额正在萎缩。官僚机构很无聊。创新发生在别处。最后一根稻草是英特尔在摩尔定律上的失误,因为该公司在其制造工艺的计划改进方面面临一系列延误,并且仍在努力纠正。自 2015 年以来,英特尔一再宣布推迟其 10nm 和 7nm 制造工艺,尽管台积电和三星已经提前推进。

As Intel approached its fiftieth anniversary in 2018, decay had set in. The company’s market share was shrinking. The bureaucracy was stultifying. Innovation happened elsewhere. The final straw was Intel’s bungling of Moore’s Law, as the company faced a series of delays to planned improvements in its manufacturing process, which it is still struggling to rectify. Since 2015, Intel has repeatedly announced delays to its 10nm and 7nm manufacturing processes, even as TSMC and Samsung have charged ahead.

公司有几乎没有解释出了什么问题。英特尔现在花了五年时间宣布“临时”制造延迟,其技术细节在员工保密协议的保密中被掩盖了。大多数业内人士认为,该公司的许多问题都源于英特尔的延迟采用 EUV 工具。到 2020 年,在英特尔资助和培育的所有 EUV 光刻工具中,有一半是安装在台积电。相比之下,英特尔才刚刚开始在其制造过程中使用 EUV。

The company has done little to explain what went wrong. Intel has now spent half a decade announcing “temporary” manufacturing delays, the technical details of which are obscured in the secrecy of employee nondisclosure agreements. Most people in the industry think many of the company’s problems stem from Intel’s delayed adoption of EUV tools. By 2020, half of all EUV lithography tools, funded and nurtured by Intel, were installed at TSMC. By contrast, Intel had only barely begun to use EUV in its manufacturing process.

随着十年的结束,只有两家公司可以制造最尖端的处理器,台积电和三星。就美国而言,两者都有问题的原因相同:它们的位置。现在,全世界先进处理器的生产都在台湾和韩国进行——就在美国新兴战略竞争对手中华人民共和国的海岸附近。

As the decade ended, only two companies could manufacture the most cutting-edge processors, TSMC and Samsung. And so far as the United States was concerned, both were problematic for the same reason: their location. Now the entire world’s production of advanced processors was taking place in Taiwan and Korea—just off the coast from America’s emerging strategic competitor: the People’s Republic of China.

第七部分 中国的挑战

PART VII CHINA’S CHALLENGE

第 42 章 中国制造

CHAPTER 42 Made in China

没有网络安全就没有国家安全,”中国共产党总书记习近平在 2014 年宣称,“没有信息化,就没有现代化。” 作为中国最早的共产党领导人之一的儿子,习近平在登上中国政坛之前曾在大学学习工程学,这要归功于他的变色龙般的本领,可以随心所欲地表现出来。对于中国民族主义者来说,他的“中国梦”计划承诺了民族复兴和大国地位。对于企业,他承诺进行经济改革。一些外国人甚至将他视为秘密民主人士,《纽约客》在他掌权后立即宣称习近平是“一位意识到中国必须进行真正的政治改革。” 唯一可以确定的是习近平作为政治家的才能。噘起的嘴唇和假装的微笑隐藏着他自己的观点。

“Without cybersecurity there is no national security,” declared Xi Jinping, general secretary of the Chinese Communist Party, in 2014, “and without informatization, there is no modernization.” The son of one of China’s earliest Communist Party leaders, Xi had studied engineering in college before ascending the ranks of Chinese politics thanks to his chameleonlike knack for appearing to be whatever a given audience thought it wanted. To Chinese nationalists, his program of a “Chinese Dream” promised national rejuvenation and great power status. To businesses, he pledged economic reform. Some foreigners even saw him as a closet democrat, with the New Yorker declaring right after he took power that Xi was “a leader who realizes that China must undertake real political reform.” The only certainty was Xi’s talent as a politician. His own views were hidden behind pursed lips and a feigned smile.

这种微笑的背后是一种刺痛的不安全感,这种不安全感在习近平统治中国的十年中推动了他的政策。他认为,主要风险是数字世界。大多数观察家认为,在保障自己的数字安全方面,习近平没什么好怕的。中国领导人拥有世界上最有效的互联网控制系统,雇用成千上万的审查员对在线聊天进行监​​管。中国的防火墙使其公民无法访问大量互联网,果断地反驳了西方关于互联网将成为自由化政治力量的预测。习近平在网上感到足够强大,足以嘲笑西方认为互联网会传播民主价值观的信念。“互联网已经把世界变成了一个地球村,”习近平宣称,回避了许多世界上最受欢迎的网站,如谷歌和 Facebook,在中国被禁止的事实。他心目中的全球网络不同于早期互联网时代的乌托邦——中国政府可以用来投射权力的网络。“我们必须走出去,深化国际互联网交流与合作,积极参与‘一带一路’建设,”他在不同场合宣称,他指的是他计划让世界融入中国建设的基础设施。不仅包括道路和桥梁,还包括网络设备和审查工具。

Behind this smile is a gnawing sense of insecurity that has driven Xi’s policies during the decade he’s ruled China. The primary risk, he believed, was the digital world. Most observers thought Xi had little to fear when it came to guaranteeing his own digital security. China’s leaders have the world’s most effective system of internet control, employing many thousands of censors to police online chatter. China’s firewall made a huge swath of the internet inaccessible to its citizens, decisively disproving Western predictions that the internet would be a liberalizing political force. Xi felt strong enough online to mock the Western belief that the internet would spread democratic values. “The internet has turned the world into a global village,” Xi declared, sidestepping the fact that many of the world’s most popular websites, like Google and Facebook, were banned in China. He had a different type of global network in mind than the utopians of the early internet age—a network that China’s government could use to project power. “We must march out, deepen international internet exchange and collaboration, and vigorously participate in the construction of ‘One Belt, One Road,’ ” he declared on a different occasion, referring to his plan to enmesh the world in Chinese-built infrastructure that included not only roads and bridges but network equipment and censorship tools.

在利用数字世界实现专制目的方面,没有哪个国家比中国更成功。它驯服了美国的科技巨头。谷歌和 Facebook 被百度和腾讯等本土公司取而代之,这些公司在技术上与美国竞争对手不相上下。赢得中国市场准入的美国科技公司,如苹果和微软,只有在同意与北京的审查工作合作后才被允许进入。中国比任何其他国家都更让互联网服从其领导人的意愿。外国互联网和软件公司要么签署了共产党想要的任何审查规则,要么失去了进入广阔市场的机会。

No country has been more successful than China at harnessing the digital world for authoritarian purposes. It has tamed America’s tech giants. Google and Facebook were banned and replaced by homegrown firms like Baidu and Tencent, which, technologically, are close matches with their American rivals. The U.S. tech firms that have won access to the Chinese market, like Apple and Microsoft, were allowed in only after agreeing to collaborate with Beijing’s censorship efforts. Far more than any other country, China has made the internet subservient to its leaders’ wishes. Foreign internet and software companies either signed on to whatever censorship rules the Communist Party desired or lost access to a vast market.

那么,为什么习近平会担心数字安全呢?中国领导人越是研究他们的技术能力,他们的互联网公司似乎就越不重要。中国的数字世界运行在数字(1 和 0)上,这些数字主要由进口半导体处理和存储。中国的科技巨头依赖于装满外国芯片(主要是美国生产的芯片)的数据中心。爱德华·斯诺登 (Edward Snowden) 在逃往俄罗斯之前于 2013 年泄露的文件展示了美国的网络窃听能力,甚至令北京的网络侦探感到惊讶。中国公司复制了硅谷的专长为电子商务、在线搜索和数字支付构建软件。但所有这些软件都依赖于外国硬件。在支撑计算的核心技术方面,中国严重依赖外国产品,其中许多产品是在硅谷设计的,而且几乎所有产品都是由美国或其盟友之一的公司生产的。

Why, then, was Xi Jinping worried about digital security? The more China’s leaders studied their technological capabilities, the less important their internet companies seemed. China’s digital world runs on digits—1s and 0s—that are processed and stored mostly by imported semiconductors. China’s tech giants depend on data centers full of foreign, largely U.S.-produced, chips. The documents that Edward Snowden leaked in 2013 before fleeing to Russia demonstrated American network-tapping capabilities that surprised even the cyber sleuths in Beijing. Chinese firms had replicated Silicon Valley’s expertise in building software for e-commerce, online search, and digital payments. But all this software relies on foreign hardware. When it comes to the core technologies that undergird computing, China is staggeringly reliant on foreign products, many of which are designed in Silicon Valley and almost all of which are produced by firms based in the U.S. or one of its allies.

习近平认为这带来了无法承受的风险。“再大的规模,再高的市值,如果一个互联网企业在核心部件上严重依赖外部世界,2016年,习近平宣称,供应链的‘要害门’掌握在别人手中。习近平最担心哪些核心技术?一个是软件产品 Microsoft Windows,尽管中国一再努力开发具有竞争力的中国操作系统,但大多数 PC 都在使用它。然而,在习近平看来,更重要的是为中国的计算机、智能手机和数据中心提供动力的芯片。正如他所指出的,“微软的 Windows 操作系统只能与英特尔芯片配对。” 所以中国的大多数计算机都需要美国芯片才能运行。在 2000 年代和 2010 年代的大部分时间里,中国花费进口半导体的钱比石油还多。在推动中国经济增长方面,高性能芯片与碳氢化合物一样重要。不过,与石油不同的是,芯片供应被中国的地缘政治对手垄断。

Xi thought this presented an untenable risk. “However great its size, however high its market capitalization, if an internet enterprise critically relies on the outside world for core components, the ‘vital gate’ of the supply chain is grasped in the hands of others,” Xi declared in 2016. Which core technologies most worry Xi? One is a software product, Microsoft Windows, which is used by most PCs in China, despite repeated efforts to develop competitive Chinese operating systems. Yet even more important in Xi’s thinking are the chips that power China’s computers, smartphones, and data centers. As he noted, “Microsoft’s Windows operating system can only be paired with Intel chips.” So most computers in China needed American chips to function. During most years of the 2000s and 2010s, China spent more money importing semiconductors than oil. High-powered chips were as important as hydrocarbons in fueling China’s economic growth. Unlike oil, though, the supply of chips is monopolized by China’s geopolitical rivals.

大多数外国人都难以理解为什么中国会感到紧张。这个国家不是建立了价值数千亿美元的大型科技公司吗?报纸头条一再宣称中国是世界领先的科技强国之一。根据谷歌中国前负责人李开复的一本被广泛讨论的书,在人工智能方面,中国是世界两大人工智能超级大国之一。北京建立了 21 世纪人工智能和威权主义的融合,最大限度地利用监控技术。但即使是追踪中国持不同政见者及其少数民族的监控系统依赖英特尔和英伟达等美国公司的芯片。中国所有最重要的技术都建立在进口硅这一脆弱的基础之上。

Most foreigners struggled to comprehend why China felt nervous. Hadn’t the country built vast tech firms worth hundreds of billions of dollars? Newspaper headlines repeatedly declared China one of the world’s leading tech powers. When it came to artificial intelligence, the country was one of the world’s two AI Superpowers, according to a widely discussed book by Kai-Fu Lee, former head of Google China. Beijing built a twenty-first-century fusion of AI and authoritarianism, maximizing use of surveillance technology. But even the surveillance systems that track China’s dissidents and its ethnic minorities rely on chips from American companies like Intel and Nvidia. All of China’s most important technology rests on a fragile foundation of imported silicon.

中国领导人不需要偏执地认为他们的国家应该在国内制造更多的芯片。这不仅仅是为了避免供应链漏洞。与其邻国一样,中国只有生产北京领导人所谓的“核心技术”——世界其他地区都离不开的产品,才能赢得更多有价值的业务。否则,中国可能会延续 iPhone 的低利润模式。数以百万计的中国人参与了手机的组装工作,但当这些设备卖给最终用户时,苹果赚了大部分钱,其余的大部分都归于每部手机中的芯片制造商。

Chinese leaders didn’t need to be paranoid to think their country should build more chips at home. It wasn’t just about avoiding supply chain vulnerability. Like its neighbors, China can only win more valuable business if it produces what Beijing’s leaders call “core technologies”—products the rest of the world can’t live without. Otherwise, China risks continuing the low-profit pattern of what has occurred with the iPhone. Millions of Chinese are involved in assembling the phones, but when the devices are sold to end users Apple makes most of the money, with much of the rest accruing to the makers of the chips inside each phone.

中国领导人面临的问题是如何转向生产世界梦寐以求的那种芯片。当日本、台湾和韩国想打入复杂和高价值的芯片行业时,他们将资金投入半导体公司,组织政府投资,同时也敦促私人银行放贷。其次,他们试图将在美国大学接受过培训并在硅谷工作的科学家和工程师引诱回家。第三,他们与外国公司建立合作伙伴关系,但要求他们转让技术或培训当地工人。第四,他们利用硅谷公司之间的竞争——以及后来的美国人和日本人之间的竞争——与外国人竞争,为自己争取到最好的交易。“我们要推动台湾的半导体产业,KT Li,在创立台积电时告诉张忠谋。习近平也想要一个,这有什么奇怪的吗?

The question for China’s leaders was how to pivot to producing the kind of chips the world coveted. When Japan, Taiwan, and South Korea wanted to break into the complex and high-value portions of the chip industry, they poured capital into their semiconductor companies, organizing government investment but also pressing private banks to lend. Second, they tried to lure home their scientists and engineers who’d been trained at U.S. universities and worked in Silicon Valley. Third, they forged partnerships with foreign firms but required them to transfer technology or train local workers. Fourth, they played foreigners off each other, taking advantage of competition between Silicon Valley firms—and, later, between Americans and Japanese—to get the best deal for themselves. “We want to promote a semiconductor industry in Taiwan,” the island’s powerful minister, K. T. Li, had told Morris Chang while founding TSMC. Was it any surprise that Xi Jinping wanted one, too?

第 43 章 “召唤进攻”

CHAPTER 43 “Call Forth the Assault”

2017 年1月,在唐纳德·特朗普就任美国总统三天前,习近平在瑞士滑雪胜地达沃斯举行的世界经济论坛上登台阐述了中国的经济愿景。当习近平承诺通过“动态的、创新驱动的增长模式”实现“双赢”时,首席执行官和亿万富翁的听众礼貌地鼓掌。“没有人会成为贸易战的赢家,”中国国家主席在毫不含糊地挖苦即将上任的美国同行时宣称。三天后,特朗普在华盛顿发表了令人震惊的好斗的就职演说,谴责“其他国家制造我们的产品,窃取我们的公司并摧毁我们的工作”。特朗普没有接受贸易,而是宣称“保护将导致巨大的繁荣和力量。”

In January 2017, Xi took the stage at the World Economic Forum in the Swiss ski resort of Davos, three days before Donald Trump’s inauguration as U.S. president, to outline China’s economic vision. As Xi promised “win-win outcomes” via a “dynamic, innovation-driven growth model,” the audience of CEOs and billionaires applauded politely. “No one will emerge as a winner in a trade war,” the Chinese president declared, in a none-too-subtle dig at his incoming American counterpart. Three days later in Washington, Trump delivered a shockingly combative inaugural address, condemning “other countries making our products, stealing our companies and destroying our jobs.” Rather than embracing trade, Trump declared that “protection will lead to great prosperity and strength.”

习的讲话是全球领导人在对商界大亨讲话时应该说的那种哗众取宠。媒体奉承他所谓的经济开放和全球化的辩护,以应对特朗普和英国退欧等民粹主义冲击。“Xi sounding rather more presidential than US president-elect,” tweeted talking-head Ian Bremmer. “习近平提出有力地捍卫全球化,”英国《金融时报》的头条新闻报道。“世界领导人发现民粹主义反抗中的达沃斯全球化希望”,《华盛顿邮报》 宣布。“国际社会正在寻找中国,”世界经济论坛主席克劳斯·施瓦布解释说。

Xi’s speech was the sort of claptrap that global leaders were supposed to say when addressing business tycoons. The media fawned over his supposed defense of economic openness and globalization against populist shocks like Trump and Brexit. “Xi sounding rather more presidential than US president-elect,” tweeted talking-head Ian Bremmer. “Xi Jinping Delivers a Robust Defence of Globalisation,” reported the lead headline in the Financial Times. “World Leaders Find Hope for Globalization in Davos Amid Populist Revolt,” the Washington Post declared. “The international community is looking to China,” explained Klaus Schwab, the chair of the World Economic Forum.

在他首次亮相达沃斯的几个月前,习近平在北京对中国科技巨头和共产党领导人的一次关于“网络安全和信息化”的会议上发表了不同的语气。在包括华为创始人任正非、阿里巴巴首席执行官马云、中国人民解放军 (PLA) 知名研究人员和大多数中国政治精英在内的听众中,习近平告诫中国要“尽快在核心技术上取得突破”。 ” 首先,“核心技术”是指半导体。习近平没有呼吁打贸易战,但他的愿景听起来也不像是贸易和平。“我们要推动强强联合,协同进攻战略要道。攻坚核心技术研发的堡垒…… 我们不仅要发动突击,还要发出集会的号召,冲进山口。” 事实证明,唐纳德·特朗普并不是唯一一个将军事隐喻与经济政策混为一谈的世界领导人。芯片行业面临着世界第二大经济体和统治它的一党制国家的有组织攻击。

Months before his Davos debut, Xi had struck a different tone in a speech to Chinese tech titans and Communist Party leaders in Beijing for a conference on “cyber security and informatization.” To an audience that included Huawei founder Ren Zhengfei, Alibaba CEO Jack Ma, high-profile People’s Liberation Army (PLA) researchers, and most of China’s political elite, Xi exhorted China to focus on “gaining breakthroughs in core technology as quickly as possible.” Above all, “core technology” meant semiconductors. Xi didn’t call for a trade war, but his vision didn’t sound like trade peace, either. “We must promote strong alliances and attack strategic passes in a coordinated manner. We must assault the fortifications of core technology research and development…. We must not only call forth the assault, we must also sound the call for assembly, which means that we must concentrate the most powerful forces to act together, compose shock brigades and special forces to storm the passes.” Donald Trump, it turned out, wasn’t the only world leader who mixed martial metaphors with economic policy. The chip industry faced an organized assault by the world’s second-largest economy and the one-party state that ruled it.

中国领导人指望结合市场和军事手段在国内开发先进芯片。尽管习近平囚禁了他的对手并成为自毛泽东以来中国最强大的领导人,但他对中国的控制远非绝对。他可以锁定持不同政见者,甚至审查网络上最隐蔽的批评。但习近平的经济议程的许多方面,从产业结构调整到金融市场改革,仍然停滞不前,受到共产党官僚和地方官员的阻碍。喜欢现状的政府官员。官员们在面对他们不喜欢的北京方面的指示时常常拖拖拉拉。

China’s leaders were counting on a mix of market and military methods to develop advanced chips at home. Though Xi had jailed his rivals and become China’s most powerful leader since Mao Zedong, his control over China was far from absolute. He could lock up dissidents and censor even the most veiled criticism online. But many facets of Xi’s economic agenda, from industrial restructuring to financial market reform, remained stillborn, obstructed by Communist Party bureaucrats and local government officials who preferred the status quo. Officials often dragged their feet when faced with instructions from Beijing that they disliked.

然而,习近平的军事言论不仅仅是动员懒惰官僚的策略。随着时间的流逝,中国技术地位的不稳定性越来越明显。中国进口的半导体逐年增加。芯片行业正在以不利于中国的方式发生变化。中国国务院在一份技术政策报告中指出:“投资规模迅速上升,市场份额加速向主导企业集中。” 这些占主导地位的公司——台积电和三星是其中的佼佼者——将很难被取代。然而,在“云计算、物联网和大数据”的推动下,中国领导人意识到,对芯片的需求正在“爆炸式增长”。这些趋势是危险的:芯片变得更加重要,但最先进芯片的设计和生产却被少数几家公司垄断,这些公司都位于中国以外。

Xi’s military rhetoric wasn’t solely a tactic for mobilizing lazy bureaucrats, however. With every year that passed, the precariousness of China’s technological position became clearer. China’s imports of semiconductors increased year after year. The chip industry was changing in ways that weren’t favorable to China. “The scale of investment has risen rapidly and market share has accelerated to the concentration of dominant firms,” China’s State Council noted in one technology policy report. These dominant firms—TSMC and Samsung chief among them—would be extremely difficult to displace. Yet demand for chips was “exploding,” China’s leaders realized, driven by “cloud computing, the Internet of Things, and big data.” These trends were dangerous: chips were becoming even more important, yet the design and production of the most advanced chips was monopolized by a handful of companies, all located outside of China.

中国的问题不仅在于芯片制造。在半导体生产过程的几乎每一步,中国都严重依赖外国技术,而这些技术几乎都被中国的地缘政治对手——台湾、日本、韩国或美国所控制。用于设计芯片的软件工具由美国公司主导,而中国在全球软件工具市场的份额不到 1%。数据由乔治城大学安全与新兴技术中心的学者汇总。在核心知识产权方面,作为设计许多芯片的晶体管模式的基石,中国的市场份额为 2%;其余大部分是美国人或英国人。中国供应全球4%的硅片和其他芯片制造材料;1% 用于制造芯片的工具;芯片设计市场的 5%。它在芯片制造业务中仅占 7% 的市场份额。这些制造能力都不涉及高价值的前沿技术。

China’s problem isn’t only in chip fabrication. In nearly every step of the process of producing semiconductors, China is staggeringly dependent on foreign technology, almost all of which is controlled by China’s geopolitical rivals—Taiwan, Japan, South Korea, or the United States. The software tools used to design chips are dominated by U.S. firms, while China has less than 1 percent of the global software tool market, according to data aggregated by scholars at Georgetown University’s Center for Security and Emerging Technology. When it comes to core intellectual property, the building blocks of transistor patterns from which many chips are designed, China’s market share is 2 percent; most of the rest is American or British. China supplies 4 percent of the world’s silicon wafers and other chipmaking materials; 1 percent of the tools used to fabricate chips; 5 percent of the market for chip designs. It has only a 7 percent market share in the business of fabricating chips. None of this fabrication capacity involves high-value, leading-edge technology.

在整个半导体供应链中,综合芯片设计、知识产权、工具、制造和其他步骤的影响,中国公司拥有 6% 的市场份额,而美国为 39%,韩国为 16%,台湾为 12%,根据乔治城研究人员的说法。几乎所有在中国生产的芯片也可以在其他地方制造。对于高级逻辑,内存,和模拟芯片,然而,中国严重依赖美国的软件和设计;美国、荷兰和日本机械;以及韩国和台湾制造。难怪习近平担心。

Across the entire semiconductor supply chain, aggregating the impact of chip design, intellectual property, tools, fabrication, and other steps, Chinese firms have a 6 percent market share, compared to America’s 39 percent, South Korea’s 16 percent, or Taiwan’s 12 percent, according to the Georgetown researchers. Almost every chip produced in China can also be fabricated elsewhere. For advanced logic, memory, and analog chips, however, China is crucially dependent on American software and designs; American, Dutch, and Japanese machinery; and South Korean and Taiwanese manufacturing. It’s no wonder that Xi Jinping was worried.

随着中国科技公司进一步进军云计算、自动驾驶汽车和人工智能等领域,它们对半导体的需求肯定会增长。仍然是现代数据中心主力的 x86 服务器芯片仍由 AMD 和英特尔主导。没有一家中国公司生产出具有商业竞争力的 GPU,让中国在这些芯片上也依赖英伟达和 AMD。正如北京的支持者所承诺的以及中国政府所希望的那样,中国越是成为人工智能超级大国,该国对外国芯片的依赖就会越多,除非中国找到设计和制造自己的芯片的方法。习近平呼吁“组成突击旅和特种部队攻关”似乎很紧迫。中国政府制定了一项名为《中国制造 2025》的计划,该计划设想将中国进口在其芯片生产中的份额从 2015 年的 85% 降至到 2025 年达到 30%。

As China’s tech firms pushed further into spheres like cloud computing, autonomous vehicles, and artificial intelligence, their demand for semiconductors was guaranteed to grow. The x86 server chips that remain the workhorse of modern data centers are still dominated by AMD and Intel. There’s no Chinese firm that produces a commercially competitive GPU, leaving China reliant on Nvidia and AMD for these chips, too. The more China becomes an AI superpower, as Beijing’s boosters promise and as China’s government hopes, the more the country’s reliance on foreign chips will increase, unless China finds a way to design and manufacture its own. Xi’s call to “compose shock brigades and special forces to storm the passes” seemed urgent. China’s government set out a plan called Made in China 2025, which envisioned reducing China’s imported share of its chip production from 85 percent in 2015 to 30 percent by 2025.

当然,自中华人民共和国成立以来,每一位中国领导人都想要一个半导体产业。毛泽东的文革梦想,即每个工人都可以生产自己的晶体管,但它彻底失败了。几十年后,中国领导人招募张国荣创立中芯国际并“与中国人分享上帝的爱”。他建立了一个有能力的代工厂,但它很难赚钱,并与台积电遭遇了一系列激烈的知识产权诉讼。最终,张被赶下台,私营部门投资者被中国政府取代。到 2015 年,中国工业和信息化部的一位前官员被任命为董事长,巩固了中芯国际与中国政府之间的关系。该公司在制造能力上继续明显落后于台积电。

Every Chinese leader since the founding of the People’s Republic wanted a semiconductor industry, of course. Mao’s Cultural Revolution dream that every worker could produce their own transistors had been an abject failure. Decades later, Chinese leaders recruited Richard Chang to found SMIC and “share God’s love with the Chinese.” He built a capable foundry, but it struggled to make money and suffered a series of bruising intellectual property lawsuits with TSMC. Eventually Chang was ousted and private-sector investors were displaced by the Chinese state. By 2015, a former official from China’s Ministry of Industry and Information was named chairman, solidifying the relationship between SMIC and the Chinese government. The firm continued to lag meaningfully behind TSMC in manufacturing prowess.

与此同时,中芯国际是中国制造行业比较成功的案例。另外两家中国代工厂华虹和格雷斯几乎没有赢得市场份额,这在很大程度上是因为国有企业控制它们的市政府不断干预商业决策。一位中国代工厂的前首席执行官解释说,每位省长都希望在他的省内建立芯片工厂,并提供各种补贴和隐蔽的威胁来确保工厂建成。因此,中国的代工厂最终以低效收集遍布全国的小型设施。外国人看到了中国芯片行业的巨大潜力,但前提是灾难性的公司治理和业务流程能够以某种方式得到解决。“当一家中国公司说,'让我们开一家合资企业吧,'”一位欧洲半导体高管解释说。“我听说,“让我们赔钱吧。” ” 确实出现的合资企业普遍沉迷于政府补贴,很少生产出有意义的新技术。

SMIC, meanwhile, was the comparative success story in China’s fabrication industry. Huahong and Grace, two other Chinese foundries, won little market share, in large part because the state-owned firms and municipal governments that controlled them meddled incessantly in business decisions. One former CEO of a Chinese foundry explained that every governor wanted a chip fab in his province and offered a mix of subsidies and veiled threats to ensure a facility was built. So China’s foundries ended up with an inefficient collection of small facilities spread across the country. Foreigners saw immense potential in the Chinese chip industry, but only if disastrous corporate governance and business processes could somehow be fixed. “When a Chinese firm said, ‘Let’s open a joint venture,’ ” one European semiconductor executive explained. “I heard, ‘Let’s lose money.’ ” The joint ventures that did emerge were generally addicted to government subsidies and rarely produced meaningful new technology.

中国 2000 年代的补贴战略并没有创造出领先的国内芯片产业。然而,什么都不做——并容忍对外国半导体的持续依赖——在政治上是不能容忍的。因此,早在 2014 年,北京就决定将半导体补贴翻倍,启动了所谓的“大基金”来支持芯片的新飞跃。该基金的主要“投资者”包括中国财政部、国有的国家开发银行以及其他各种国有企业,包括中国烟草和北京、上海和武汉市政府的投资机构。一些分析师将此称为国家支持的新型“风险投资”模式,但强制中国国有卷烟公司投资集成电路的决定,与硅谷风险投资的运营模式相差无几。

China’s subsidy strategy of the 2000s hadn’t created a leading-edge domestic chip industry. Yet doing nothing—and tolerating continued dependence on foreign semiconductors—wasn’t politically tolerable. So as early as 2014, Beijing had decided to double down on semiconductor subsidies, launching what became known as the “Big Fund” to back a new leap forward in chips. Key “investors” in the fund include China’s Ministry of Finance, the state-owned China Development Bank, and a variety of other government-owned firms, including China Tobacco and investment vehicles of the Beijing, Shanghai, and Wuhan municipal governments. Some analysts hailed this as a new “venture capital” model of state support, but the decision to force China’s state-owned cigarette company to fund integrated circuits was about as far from the operating model of Silicon Valley venture capital as could be.

北京的结论是正确的,该国的芯片产业需要更多资金。2014 年,当该基金推出时,先进晶圆厂的成本远远超过 100 亿美元。中芯国际报告称,整个 2010 年代的年收入仅为数十亿美元,不到台积电的十分之一。仅靠私营部门的资金来复制台积电的投资计划是不可能的。只有政府才能进行这样的赌博。中国投入芯片补贴和“投资”的金额难以计算,因为大部分支出已经完成由地方政府和不透明的国有银行,但人们普遍认为是数百亿美元。

Beijing was right to conclude the country’s chip industry needed more money. In 2014, when the fund was launched, advanced fabs cost well over $10 billion. SMIC reported revenue of just a couple billion dollars per year throughout the 2010s, less than a tenth of TSMC. It would be impossible to replicate TSMC’s investment plans with private-sector funding alone. Only a government could take such a gamble. The amount of money China’s put into chip subsidies and “investments” is hard to calculate, since much of the spending is done by local governments and opaque state-owned banks, but it’s widely thought to measure in the tens of billions of dollars.

然而,中国处于不利地位,因为政府不想与硅谷建立联系,而是想摆脱它。日本、韩国、荷兰和台湾通过与美国芯片产业深度融合,在半导体生产过程的重要步骤中占据主导地位。台湾的晶圆代工业之所以富足,是因为美国的无晶圆厂公司,而 ASML 最先进的光刻工具,也只有在该公司圣地亚哥子公司生产的专业光源下才能发挥作用。尽管在贸易方面偶尔会出现紧张局势,但这些国家有着相似的利益和世界观,因此在芯片设计、工具和制造服务方面相互依赖被视为为提高全球化生产效率付出的合理代价。

China was disadvantaged, however, by the government’s desire not to build connections with Silicon Valley, but to break free of it. Japan, South Korea, the Netherlands, and Taiwan had come to dominate important steps of the semiconductor production process by integrating deeply with the U.S. chip industry. Taiwan’s foundry industry only grew rich thanks to America’s fabless firms, while ASML’s most advanced lithography tools only work thanks to specialized light sources produced at the company’s San Diego subsidiary. Despite occasional tension over trade, these countries have similar interests and worldviews, so mutual reliance on each other for chip designs, tools, and fabrication services was seen as a reasonable price to pay for the efficiency of globalized production.

如果中国只想在这个生态系统中占据更大的份额,它的野心本可以得到满足。然而,北京并没有在由美国及其朋友主导的体系中寻找更好的位置。习近平“攻城略地”的号召并不是要稍微提高市场份额。这是关于重塑世界半导体产业,而不是与之整合。中国的一些经济政策制定者和半导体行业高管更倾向于更深层次的整合战略,但北京的领导人更多地考虑安全而不是效率,将相互依存视为一种威胁。《中国制造2025》不是提倡经济一体化,而是相反。它呼吁削减中国对进口芯片的依赖。中国制造2025计划的主要目标是减少在中国使用的外国芯片的份额。

If China only wanted a bigger part in this ecosystem, its ambitions could’ve been accommodated. However, Beijing wasn’t looking for a better position in a system dominated by America and its friends. Xi’s call to “assault the fortifications” wasn’t a request for slightly higher market share. It was about remaking the world’s semiconductor industry, not integrating with it. Some economic policymakers and semiconductor industry executives in China would have preferred a strategy of deeper integration, yet leaders in Beijing, who thought more about security than efficiency, saw interdependence as a threat. The Made in China 2025 plan didn’t advocate economic integration but the opposite. It called for slashing China’s dependence on imported chips. The primary target of the Made in China 2025 plan is to reduce the share of foreign chips used in China.

这种经济愿景有可能改变贸易流动和全球经济。自从飞兆半导体在香港开设第一家工厂以来,芯片贸易就帮助建立了全球化。中国重塑半导体供应链的愿景所涉及的美元价值令人震惊。中国芯片进口——2017 年为 2600 亿美元习近平在达沃斯的首次亮相——远远超过沙特阿拉伯的石油出口或德国的汽车出口。中国每年花在购买芯片上的钱比整个全球的飞机贸易还多。在国际贸易中,没有任何产品比半导体更重要。

This economic vision threatened to transform trade flows and the global economy. Since Fairchild Semiconductor’s first facility in Hong Kong, trade in chips had helped build globalization. The dollar values at stake in China’s vision of reworking semiconductor supply chains were staggering. China’s import of chips—$260 billion in 2017, the year of Xi’s Davos debut—was far larger than Saudi Arabia’s export of oil or Germany’s export of cars. China spends more money buying chips each year than the entire global trade in aircraft. No product is more central to international trade than semiconductors.

面临风险的不仅是硅谷的利润。如果中国在半导体领域实现自给自足的努力取得成功,其邻国(其中大部分经济体依赖出口)将遭受更大的损失。2017年集成电路占韩国出口的15%;新加坡的 17%;马来西亚的 19%;21% 的菲律宾”;和台湾的 36%。《中国制造2025》对这一切提出了质疑。危在旦夕的是世界上最密集的供应链和贸易流动网络,以及在过去半个世纪里支撑亚洲经济增长和政治稳定的电子行业。

It wasn’t only Silicon Valley’s profits that were at risk. If China’s drive for self-sufficiency in semiconductors succeeded, its neighbors, most of whom had export-dependent economies, would suffer even more. Integrated circuits made up 15 percent of South Korea’s exports in 2017; 17 percent of Singapore’s; 19 percent of Malaysia’s; 21 percent of the Philippines’; and 36 percent of Taiwan’s. Made in China 2025 called all this into question. At stake was the world’s most dense network of supply chains and trade flows, the electronics industries that had undergirded Asia’s economic growth and political stability over the past half century.

当然,中国制造 2025 只是一个计划。政府的计划常常以失败告终。中国在刺激尖端芯片生产方面的记录远非令人印象深刻。然而,中国可以使用的工具——巨额政府补贴、国家支持的商业机密盗窃,以及利用进入世界第二大消费市场的机会迫使外国公司服从其命令的能力——赋予了北京无与伦比的塑造权力芯片行业的未来。如果有哪个国家能够实现如此雄心勃勃的贸易流动转型,那就是中国。该地区的许多国家认为北京可能会成功。台湾科技业开始担心台湾人所说的“红色供应链”——进军高价值电子元件的大陆公司此前一直占据主导地位。很容易想象半导体会是下一个。

Made in China 2025 was just a plan, of course. Governments often have plans that fail abjectly. China’s track record in spurring production of cutting-edge chips was far from impressive. Yet the tools China could bring to bear—vast government subsidies, state-backed theft of trade secrets, and the ability to use access to the world’s second-largest consumer market to force foreign firms to follow its writ—gave Beijing unparalleled power to shape the future of the chip industry. If any country could pull off such an ambitious transformation of trade flows, it was China. Many countries in the region thought Beijing might succeed. Taiwan’s tech industry began worrying about what Taiwanese called the “red supply chain”—the mainland firms muscling into high-value electronics components Taiwan had previously dominated. It was easy to imagine semiconductors would be next.

习近平呼吁中国政府及其企业“攻打核心技术研究的堡垒”,早在对西方产生重大影响之前,就已经在东亚引起了反响。唐纳德特朗普关于保护主义的声明获得了数百万次转发,但北京有一个计划、强大的工具和四十年的轨道以中国的经济和技术实力震惊世界的记录。这种半导体独立的愿景有望颠覆全球化,改变世界上交易最广泛和最有价值的商品之一的生产方式。2017 年习近平在达沃斯发表演讲的听众中没有人注意到陈词滥调背后的利害关系,但即使是像特朗普这样的民粹主义者也无法想象全球经济会发生更激进的改造。

Xi Jinping’s call for China’s government and its companies to “assault the fortifications of core technology research” reverberated around East Asia long before it made much impact in the West. Donald Trump’s proclamations about protectionism garnered millions of retweets, but Beijing had a plan, powerful tools, and a forty-year track record of surprising the world with China’s economic and technological capabilities. This vision of semiconductor independence promised to upend globalization, transforming the production of one of the world’s most widely traded and most valuable goods. No one in the audience of Xi’s speech at Davos in 2017 noticed what was at stake behind the platitudes, but even a populist like Trump couldn’t have imagined a more radical reworking of the global economy.

第 44 章 技术转让

CHAPTER 44 Technology Transfer

如果你是一个拥有 13 亿人口的国家,就像中国一样,你会想要一个 IT 产业,”IBM 首席执行官 Ginni Rometty 在中国政府在北京举办的年度活动 2015 中国发展论坛上对听众说。“我认为一些公司发现这可能令人恐惧。但是,在 IBM,我们……发现这是一个很好的机会。” 在所有美国科技公司中,没有一家比 IBM 与美国政府的关系更密切。近一个世纪以来,该公司为美国最敏感的国家安全应用构建了先进的计算机系统。IBM 员工与五角大楼和美国情报机构的官员有着深厚的私人关系。当爱德华·斯诺登 (Edward Snowden) 在逃往莫斯科之前窃取并泄露了有关美国对外情报行动的文件时,发现 IBM 因涉嫌与美国网络侦探合作。

“If you’re a country, as China is, of 1.3 billion people, you would want an IT industry,” IBM CEO Ginni Rometty told the audience at the 2015 China Development Forum, an annual event hosted by China’s government in Beijing. “I think some firms find that perhaps frightening. We, though, at IBM… find that to be a great opportunity.” Of all America’s tech firms, none had a closer relationship to the U.S. government than IBM. For nearly a century, the company had built advanced computer systems for America’s most sensitive national security applications. IBM staff had deep personal relationships with officials in the Pentagon and in U.S. intelligence agencies. When Edward Snowden stole and leaked documents about America’s foreign intelligence operations before fleeing to Moscow, it wasn’t a surprise to find IBM under suspicion for collaborating with American cyber sleuths.

斯诺登泄密事件后,IBM 在中国的销售额下降了 20%,因为中国公司转向其他地方购买服务器和网络设备。IBM 首席财务官 Martin Schroeter 告诉投资者,“中国正在经历一套非常重要的经济改革”,这是一种雄辩的方式来解释中国政府通过限制其销售来惩罚 IBM。罗睿兰决定以半导体技术的形式向北京伸出橄榄枝。2014年后的几年里,她对中国进行了一系列访问,与中国高级官员会面,例如李克强总理、北京市市长王安顺和亲自负责中国芯片产业升级的副总理马凯。IBM 告诉媒体,罗睿兰此次访问北京的目的是“强调这家科技巨头对当地伙伴关系、未来合作和信息安全的承诺”,因为路透社的一篇报道称。中国国营的新华社对交换条件更加直言不讳,报道称罗梅蒂和马云讨论过“加强集成电路合作”发展。

After the Snowden leaks, IBM’s sales in China slumped by 20 percent as Chinese firms turned elsewhere for servers and networking equipment. IBM’s CFO, Martin Schroeter, told investors that “China is going through a very significant economic set of reforms,” an eloquent way of explaining that the Chinese government was punishing IBM by restricting its sales. Rometty decided to offer Beijing an olive branch in the form of semiconductor technology. She made a series of visits to China in the years after 2014, meeting with top Chinese officials like Premier Li Keqiang, Beijing mayor Wang Anshun, and Vice Premier Ma Kai, who was personally in charge of China’s efforts to upgrade its chip industry. IBM told the media that Rometty’s visits to Beijing were intended “to emphasize the tech giant’s commitment to local partnerships, future cooperation, and information security,” as a report by the Reuters news agency put it. China’s state-run Xinhua news service was even more blunt about the quid pro quo, reporting that Rometty and Ma discussed “enhancing cooperation in integrated circuit” development.

在推动半导体自给自足的过程中,北京的重点领域之一是服务器芯片。2010 年代中期非常像今天,世界数据中心主要依赖使用 x86 指令集架构的芯片,尽管 Nvidia 的 GPU 开始赢得市场份额。只有三家公司拥有生产 x86 芯片所需的知识产权:美国的英特尔和 AMD 以及一家名为 Via 的台湾小公司。实际上,英特尔主导了市场。IBM 的“Power”芯片架构曾经在企业服务器中扮演过重要角色,但在 2010 年代就失败了。一些研究人员认为,在移动设备中流行的 Arm 架构也可能在未来的数据中心中发挥作用,尽管当时基于 Arm 的芯片已经服务器市场份额很小。无论采用何种架构,中国在国内几乎没有能力生产具有竞争力的数据中心芯片。中国政府着手获取这项技术,对美国公司进行强力武装,并迫使它们将技术转让给中国合作伙伴。

In its drive for semiconductor self-sufficiency, one of Beijing’s focus areas was chips for servers. The mid-2010s were very much like today where the world’s data centers rely mostly on chips using the x86 instruction set architecture, though Nvidia’s GPUs were beginning to win market share. Only three companies had the necessary intellectual property to produce x86 chips: America’s Intel and AMD as well as a small Taiwanese company called Via. In practice, Intel dominated the market. IBM’s “Power” chip architecture had once played a major role in corporate servers but had lost out in the 2010s. Some researchers thought that Arm’s architecture—popular in mobile devices—might also play a role in future data centers, though at the time Arm-based chips had little server market share. Whatever the architecture, China had virtually no domestic capability to produce competitive data center chips. China’s government set out to acquire this technology, strong-arming U.S. companies and pressuring them to transfer technology to Chinese partners.

主导服务器半导体销售的英特尔几乎没有动力与北京就数据中心处理器达成交易(尽管它在移动芯片和 NAND 存储芯片市场分别与中国国有企业和地方政府进行交易,其中英特尔的地位较弱)。然而,在数据中心市场份额被英特尔抢走的美国芯片制造商正在寻找竞争优势。在 IBM,罗密蒂宣布了一项吸引北京的战略改变。她宣布,IBM 不会试图向中国客户销售芯片和服务器,而是开放其她解释说,向中国合作伙伴提供芯片技术,使他们能够“创建一个新的、充满活力的中国公司生态系统,为本地和国际市场生产本土计算机系统。” IBM 以技术换取市场准入的决定具有商业意义。该公司的技术被认为是二流的,没有北京的认可,它不太可能扭转其在斯诺登之后的市场萎缩。IBM 同时试图将其全球业务从销售硬件转向销售服务,因此共享对其芯片设计的访问权限似乎是合乎逻辑的。

Intel, which dominated sales of semiconductors for servers, had few incentives to cut deals with Beijing over data center processors (though it was separately doing deals with Chinese state-backed firms and local governments in the market for mobile chips and NAND memory chips, where Intel’s position was weaker). The American chipmakers that had lost data center market share to Intel, however, were looking for a competitive advantage. At IBM, Rometty announced a change of strategy that would appeal to Beijing. Rather than trying to sell chips and servers to Chinese customers, she announced, IBM would open its chip technology to Chinese partners, enabling them, she explained, to “create a new and vibrant ecosystem of Chinese companies producing homegrown computer systems for the local and international markets.” IBM’s decision to trade technology for market access made business sense. The firm’s technology was seen as second-rate, and without Beijing’s imprimatur it was unlikely to reverse its post-Snowden market shrinkage. IBM was simultaneously trying to shift its global business from selling hardware to selling services, so sharing access to its chip designs seemed logical.

然而,对于中国政府来说,这种伙伴关系不仅仅关乎商业。据《纽约时报》报道,与 IBM 新推出的芯片技术合作的人之一是中国核导弹库前网络安全负责人沉长祥。就在一年前,沉曾警告过与美国公司合作存在“巨大的安全风险”。现在他似乎已经得出结论,IBM 提供芯片技术的提议支持北京的半导体战略和中国的国家利益。

For China’s government, however, this partnership wasn’t solely about business. One of the individuals working with IBM’s newly available chip technology was the former cyber security chief of China’s nuclear missile arsenal, Shen Changxiang, the New York Times reported. Just a year earlier, Shen had been warning of the “huge security risks” in working with U.S. firms. Now he appeared to have concluded that IBM’s offer to turn over chip technology supported Beijing’s semiconductor strategy and China’s national interests.

IBM 并不是唯一一家愿意帮助中国公司开发数据中心芯片的公司。大约在同一时间,专注于智能手机芯片的公司高通正试图使用​​ Arm 架构打入数据中心芯片业务。与此同时,高通正在与中国监管机构作斗争,中国监管机构希望它削减向授权其智能手机芯片技术的中国公司收取的费用,高通收入的主要来源。作为高通芯片的最大市场,中国对该公司拥有巨大的影响力。因此,在与北京解决价格纠纷后不久,高通同意与一家名为华芯通的中国公司合资开发服务器芯片,因此一些行业分析师看到了这种联系。行业分析人士指出,华芯通在先进芯片设计方面没有业绩记录,但它位于贵州省,当时由一位名叫陈敏儿的新晋中共官员领导。

IBM wasn’t the only company willing to help Chinese firms develop data center chips. Around the same time, Qualcomm, the company specializing in chips for smartphones, was trying to break into the data center chip business using an Arm architecture. Simultaneously, Qualcomm was battling Chinese regulators who wanted it to slash the fees it charged Chinese firms that licensed its smartphone chip technology, a key source of Qualcomm’s revenue. As the biggest market for Qualcomm’s chips, China had enormous leverage over the company. So some industry analysts saw a connection when, shortly after settling the pricing dispute with Beijing, Qualcomm agreed to a joint venture with a Chinese company called Huaxintong to develop server chips. Huaxintong didn’t have a track record in advanced chip design, but it was based in Guizhou Province, then governed by an up-and-coming Chinese party official named Chen Min’er, industry analysts noted.

高通与华芯通的合资企业并没有持续多久。它在没有产生任何价值后于 2019 年关闭。但有些专业开发似乎已经转移到其他构建基于 Arm 的数据中心芯片的中国公司。例如,华芯通参与了一个财团开发的节能芯片,包括 Phytium,这是另一家制造基于 Arm 的芯片的中国公司。至少有一名芯片设计工程师似乎在 2019 年离开了华芯通为 Phytium 工作,美国后来称其帮助中国军方设计了高超音速导弹等先进武器系统。

The Qualcomm-Huaxintong joint venture didn’t last long. It was closed in 2019 after producing little of value. But some of the expertise developed appears to have transferred to other Chinese companies building Arm-based data center chips. For example, Huaxintong participated in a consortium to develop energy-efficient chips that included Phytium, another Chinese firm building Arm-based chips. At least one chip design engineer appears to have left Huaxintong in 2019 to work for Phytium, which the U.S. later alleged had helped the Chinese military design advanced weapons systems like hypersonic missiles.

然而,最具争议的技术转让例子是英特尔的主要竞争对手 AMD。在 2010 年代中期,该公司陷入财务困境,PC 和数据中心市场份额被英特尔抢走。AMD 从未处于破产边缘,但也离破产不远。该公司在将新产品推向市场时,一直在寻找现金来争取时间。例如,2013 年,它出售了位于德克萨斯州奥斯汀的公司总部,以筹集现金。2016 年,它以 3.71 亿美元的价格将其在马来西亚槟城和中国苏州的半导体组装、测试和封装设施的 85% 股权出售给了一家中国公司。AMD 将这些设施描述为“世界级。”

The most controversial example of technology transfer, however, was by Intel’s archrival, AMD. In the mid-2010s, the company was struggling financially, having lost PC and data center market share to Intel. AMD was never on the brink of bankruptcy, but it wasn’t far from it, either. The company was looking for cash to buy time as it brought new products to market. In 2013, it sold its corporate headquarters in Austin, Texas, to raise cash, for example. In 2016, it sold to a Chinese firm an 85 percent stake in its semiconductor assembly, testing, and packaging facilities in Penang, Malaysia, and Suzhou, China, for $371 million. AMD described these facilities as “world-class.”

同年,AMD 与一个财团达成协议中国公司和政府机构许可为中国市场生产改进的 x86 芯片。该交易在业内和华盛顿都引起了极大的争议,其结构无需获得美国政府委员会 CFIUS 的批准,该委员会是审查外国购买美国资产的美国政府委员会。AMD 将这笔交易提交给了商务部的相关部门,但他们没有正如一位业内人士所说,“对微处理器、半导体或中国一无所知”。据报道,英特尔就这笔交易向政府发出警告,暗示它损害了美国的利益,并将威胁到英特尔的业务。然而,政府缺乏直接的方法来阻止它,因此该交易最终被挥手通过,引发了国会和五角大楼的愤怒。

That same year, AMD cut a deal with a consortium of Chinese firms and government bodies to license the production of modified x86 chips for the Chinese market. The deal, which was deeply controversial within the industry and in Washington, was structured in a way that didn’t require the approval of CFIUS, the U.S. government committee that reviews foreign purchases of American assets. AMD took the transaction to the relevant authorities in the Commerce Department, who don’t “know anything about microprocessors, or semiconductors, or China,” as one industry insider put it. Intel reportedly warned the government about the deal, implying that it harmed U.S. interests and that it would threaten Intel’s business. Yet the government lacked a straightforward way to stop it, so the deal was ultimately waved through, sparking anger in Congress and in the Pentagon.

就在 AMD 完成交易之际,其名为“Zen”的新处理器系列开始上市,扭转了公司的命运,因此 AMD最终不依赖于其许可交易的资金。但是,合资企业已经签约,技术已经转让。《华尔街日报》刊登了多篇报道,称 AMD 出售了“皇冠上的珠宝”和“王国的钥匙”。其他行业分析师表示,该交易旨在让中国公司向中国政府声称他们正在中国设计尖端微处理器,而实际上他们是只需调整 AMD 设计。这笔交易在英文媒体上被描述为一项小型许可交易,但中国领先的专家告诉国有媒体,这笔交易支持中国本土化“核心技术”的努力,以便“我们不再被我们的鼻子牵着鼻子走”。反对这笔交易的五角大楼官员同意 AMD 严格遵守法律条文,但表示他们仍然不相信这笔交易像捍卫者声称的那样无害。“我仍然非常怀疑我们是否从 AMD 那里得到了完整的故事,”一位前五角大楼官员说。华尔街日报》报道称,合资企业涉及中国超级计算机公司曙光,该公司将“为中国的国防和安全做出贡献”描述为“根本使命”。AMD 在 2017 年的新闻稿中将曙光描述为“战略合作伙伴”,并保证在华盛顿引起人们的注意。

Just as AMD finalized the deal, its new processor series, called “Zen,” began hitting the market, turning around the company’s fortunes, so AMD ended up not depending on the money from its licensing deal. However, the joint venture had already been signed and the technology was transferred. The Wall Street Journal ran multiple stories arguing that AMD had sold “crown jewels” and “the keys to the kingdom.” Other industry analysts suggested the transaction was designed to let Chinese firms claim to the Chinese government they were designing cutting-edge microprocessors in China, when in reality they were simply tweaking AMD designs. The transaction was portrayed in English-language media as a minor licensing deal, but leading Chinese experts told state-owned media the deal supported China’s effort to domesticate “core technologies” so that “we no longer can be pulled around by our noses.” Pentagon officials who opposed the deal agree that AMD scrupulously followed the letter of the law, but say they remain unconvinced the transaction was as innocuous as defenders claim. “I continue to be very skeptical we were getting the full story from AMD,” one former Pentagon official says. The Wall Street Journal reported that the joint venture involved Sugon, a Chinese supercomputer firm that has described “making contributions to China’s national defense and security” as its “fundamental mission.” AMD described Sugon as a “strategic partner” in press releases as recently as 2017, which was guaranteed to raise eyebrows in Washington.

很明显,曙光希望帮助建立一些世界领先的超级计算机,这些超级计算机通常用于开发正如商务部长吉娜·雷蒙多(Gina Raimondo)在 2021 年解释的那样,“核武器和高超音速武器”。曙光本身拥有据美国著名的中国军事专家艾尔莎·卡尼亚 (Elsa Kania) 称,该公司在广告中宣传其与中国军方的联系。即使在特朗普政府决定将曙光列入黑名单,切断与 AMD 的关系后,芯片行业分析师安东·希洛夫 (Anton Shilov) 仍发现曙光电路板采用 AMD 芯片,它本不应购买。AMD 告诉记者,它没有为有问题的设备提供技术支持,并且不确定中科曙光是如何获得这些芯片的。

What’s clear is that Sugon wanted help to build some of the world’s leading supercomputers, which are commonly used for developing “nuclear weapons and hypersonic weapons,” as Commerce Secretary Gina Raimondo explained in 2021. Sugon itself has advertised its links to the Chinese military, according to Elsa Kania, a leading American expert on the Chinese military. Even after the Trump administration decided to blacklist Sugon, severing the relationship with AMD, chip industry analyst Anton Shilov found Sugon circuit boards with AMD chips that it shouldn’t have been able to buy. AMD told journalists it had not provided technical support for the device in question and wasn’t sure how Sugon acquired the chips.

中国市场如此诱人,以至于公司发现几乎不可能避免技术转让。一些公司被甚至被诱导转让其整个中国子公司的控制权。2018 年,设计芯片架构的英国公司 Arm 剥离了其中国部门,将 Arm 中国 51% 的股份出售给了一群投资者,同时保留了另外 49% 的股份。两年前,Arm 被日本公司软银收购,该公司已向中国科技初创公司投资数十亿美元。因此,软银的投资成功依赖于有利的中国监管待遇。它面临着美国监管机构的审查,他们担心它在中国的风险使其成为容易受到来自北京的政治压力。软银在 2016 年以 400 亿美元收购了 Arm,但它出售了中国部门 51% 的股份——据软银称,该部门占 Arm 全球销售额的五分之一——只有7.75亿美元。

The Chinese market was so enticing that companies found it nearly impossible to avoid transferring technology. Some companies were even induced to transfer control of their entire China subsidiaries. In 2018, Arm, the British company that designs the chip architecture, spun out its China division, selling 51 percent of Arm China to a group of investors, while retaining the other 49 percent itself. Two years earlier, Arm had been purchased by Softbank, a Japanese company that has invested billions in Chinese tech startups. Softbank was therefore dependent on favorable Chinese regulatory treatment for the success of its investments. It faced scrutiny from U.S. regulators, who worried that its exposure to China made it vulnerable to political pressure from Beijing. Softbank had purchased Arm in 2016 for $40 billion, but it sold a 51 percent stake in the China division—which according to Softbank accounted for a fifth of Arm’s global sales—for only $775 million.

拆分 Arm 中国的逻辑是什么?没有确凿的证据表明软银面临中国官员出售该公司中国子公司的压力。然而,Arm 高管在描述逻辑时持开放态度。“如果有人正在为中国军方或中国监视构建[芯片上系统],”Arm 的一位高管告诉日经亚洲,“中国只想在中国境内拥有它。有了这种新的合资企业,这家公司就可以发展它。在过去,这是我们无法做到的。” “中国想要安全可控,”这位高管继续说道。“最终,他们想要控制自己的技术…… 如果它基于我们带来的技术,我们可以从中受益,”他解释道。监管软银的日本官员、监管 Arm 的英国官员,以及对 Arm 大部分知识产权拥有管辖权的美国官员都没有选择调查其影响。

What was the logic of spinning off Arm China? There’s no hard evidence that Softbank faced pressure from Chinese officials to sell the company’s Chinese subsidiary. Arm executives were open, however, in describing the logic. “If somebody was building [a system on a chip] for China military or China surveillance,” one Arm executive told Nikkei Asia, “China wants to have it only inside China. With this kind of new joint venture, this company can develop that. In the past this is something we couldn’t do.” “China wants to be secure and controllable,” this executive continued. “Ultimately they want to have control of their technology…. If it’s based on the technology that we bring, we could benefit from that,” he explained. Neither the Japanese officials who regulate Softbank, the UK officials who regulate Arm, nor the American officials with jurisdiction over a substantial portion of Arm’s intellectual property chose to investigate the implications.

芯片公司根本不能忽视全球最大的半导体市场。当然,芯片制造商会小心翼翼地保护他们的关键技术。但几乎每家芯片公司都有非核心技术,在他们不领先的子领域,他们很乐意以一定的价格分享这些技术。此外,当公司正在失去市场份额或需要融资时,他们没有精力专注于长期。这为中国提供了强大的杠杆来吸引外国芯片公司转让技术,开放生产设施或许可知识产权,即使外国公司意识到他们正在帮助发展竞争对手。对于芯片公司来说,在中国筹集资金往往比在华尔街更容易。接受中国资本可能是在该国开展业务的一项隐含要求。

Chip firms simply can’t ignore the world’s largest market for semiconductors. Chipmakers jealously guard their critical technologies, of course. But almost every chip firm has non-core technology, in subsectors that they don’t lead, that they’d be happy to share for a price. When companies are losing market share or in need of financing, moreover, they don’t have the luxury of focusing on the long term. This gives China powerful levers to induce foreign chip firms to transfer technology, open production facilities, or license intellectual property, even when foreign companies realize they’re helping develop competitors. For chip firms, its often easier to raise funds in China than on Wall Street. Accepting Chinese capital can be an implicit requirement for doing business in the country.

从他们自己的角度来看,IBM、AMD 和 Arm 在中国达成的交易是由合理的商业逻辑驱动的。总的来说,他们冒着技术泄露的风险。美国和英国的芯片架构和设计以及台湾代工厂在中国超级计算机项目的发展中发挥了核心作用。与十年前相比,尽管其能力仍然明显落后于前沿,但中国在设计和生产数据中心所需芯片方面对外国人的依赖程度已大大降低。IBM 首席执行官 Ginni Rometty 在与中国的技术转让协议中感觉到“巨大机遇”是正确的。她认为她的公司将成为受益人是错误的。

Viewed on their own terms, the deals that IBM, AMD, and Arm struck in China were driven by reasonable business logic. Collectively, they risk technology leakage. U.S. and UK chip architectures and designs as well as Taiwanese foundries have played a central role in the development of China’s supercomputer programs. Compared to a decade ago, though its capabilities still meaningfully lag the cutting edge, China is substantially less reliant on foreigners to design and produce chips needed in data centers. IBM CEO Ginni Rometty was right to sense “great opportunity” in technology transfer agreements with China. She was only wrong in thinking her firm would be the beneficiary.

第 45 章 “并购势在必行”

CHAPTER 45 “Mergers Are Bound to Happen”

还是赵卫国,从小时候在中国西部边疆养猪养羊,到现在是一条漫长曲折的路被中国媒体誉为芯片亿万富翁。在他的父亲在文革期间因写颠覆性诗歌而被流放后,赵最终来到了中国农村,但他从未想过要接受在农村饲养牲畜的生活。他考入了中国最好的大学之一清华大学,并攻读了电气工程学位。清华大学在半导体行业最早进入中国时就一直领导着中国的半导体工作,但尚不清楚赵在学生时期培养了多少晶体管和电容器方面的专业知识。本科毕业后在一家科技公司工作,后转投投资,担任紫光集团副总裁。这家公司是他的母校成立的,旨在将大学的科研成果转化为有利可图的业务,但它似乎在房地产上投入了大量资金。赵建立了企业交易撮合者的声誉,并立志于通往十亿美元财富的道路。

For Zhao Weiguo, it was a long, winding road from a childhood raising pigs and sheep along China’s western frontier to being celebrated as a chip billionaire by Chinese media. Zhao ended up in rural China after his father was banished for writing subversive poems during the Cultural Revolution, but he never planned to accept a life rearing livestock in the countryside. He won entrance to Tsinghua University, one of the best in China, and pursued a degree in electrical engineering. Tsinghua had led China’s semiconductor efforts since the industry’s earliest days in China, but it isn’t clear how much expertise in transistors and capacitors Zhao developed as a student. He worked at a tech firm after finishing his bachelor’s degree, then pivoted toward investing as a vice president of Tsinghua Unigroup. This company was established by his alma mater to turn the university’s scientific research into profitable businesses, but it appears to have invested heavily in real estate. Zhao built a reputation as a corporate dealmaker and set himself on a path toward a billion-dollar fortune.

2004年,赵成立了自己的投资基金——北京建坤集团,投资于房地产、矿业和其他高层政治关系通常对成功至关重要的行业。财力雄厚回报随之而来,据报道,赵将100万元的初始投资资金变成了45亿元。2009年,赵利用这笔财富购买了其前东家紫光集团49%的股份。该大学继续拥有其他 51% 的股份。这是一笔奇怪的交易:一家私人房地产投资公司现在拥​​有一家公司近一半的股份,这家公司本应将中国一流研究型大学的技术货币化。但清华紫光绝不仅仅是一家“普通”的公司。中国前国家主席胡锦涛的儿子——据说是赵氏的“私人朋友”——曾担任紫光集团旗下控股公司的党委书记。与此同时,2000 年代清华大学的校长是一位习近平的大学室友。

In 2004, Zhao launched his own investment fund, Beijing Jiankun Group, investing in real estate, mining, and other sectors where high-level political connections are usually crucial to success. Rich financial returns followed, with Zhao reportedly turning 1 million yuan of initial invested capital into 4.5 billion. In 2009, Zhao used this wealth to buy a 49 percent stake in his former employer, Tsinghua Unigroup. The university continued to own the other 51 percent of shares. It was a bizarre transaction: a private real-estate investment firm now owned nearly half of a company that was supposed to be monetizing technologies produced by China’s premier research university. But Tsinghua Unigroup was never simply a “normal” company. The son of former Chinese president Hu Jintao—said to be a “personal friend” of Zhao’s—served as Communist Party secretary for the holding company that owned Unigroup. The president of Tsinghua University throughout the 2000s, meanwhile, was a college roommate of Xi Jinping.

2013 年,在收购紫光集团股份四年后,就在中国共产党宣布向中国半导体公司提供巨额补贴的新计划之前,赵决定是时候投资芯片行业了。他否认紫光集团的半导体战略是对政府意愿的回应。“每个人都认为政府正在推动芯片行业的发展,但事实并非如此,”他在 2015 年告诉《福布斯》。相反,他认为吸引了北京对该行业的关注。“公司首先做了一些事情,然后政府开始注意到......我们所有的交易都是以市场为导向的。”

In 2013, four years after buying his stake in Tsinghua Unigroup, and just before China’s Communist Party announced new plans to provide vast subsidies to the country’s semiconductor firms, Zhao decided it was time to invest in the chip industry. He denies that Tsinghua Unigroup’s semiconductor strategy was a response to the government’s wishes. “Everyone thinks that the government is pushing the development of the chip sector, but it’s not like that,” he told Forbes in 2015. Instead, he takes credit for attracting Beijing’s attention to the sector. “Companies did some stuff first and then the government started to notice…. All our deals are market oriented.”

“以市场为导向”并不是大多数分析师对赵的战略的描述。他没有投资最好的芯片公司,而是尝试在市场上购买任何东西。他对清华投资策略的解释并没有暗示细微差别或复杂性。“如果你带着枪上山,你就是不知道那里是否有游戏,”他被引述说。“也许你会抓到一只鹿,也许是一只山羊,你只是不知道。” 尽管如此,他还是一个自信的猎人。世界各地的芯片公司都是他的猎物。

“Market oriented” is not how most analysts would describe Zhao’s strategy. Rather than investing in the best chip firms, he tried buying anything on the market. His explanation of Tsinghua’s investment strategy didn’t suggest nuance or sophistication. “If you carry your gun up the mountain, you just don’t know if there’s game there,” he was quoted as saying. “Maybe you’ll catch a deer, maybe a goat, you just don’t know.” Nevertheless, he was a confident hunter. The world’s chip firms were his prey.

即使考虑到他估计为 20 亿美元的财富,赵为建立自己的芯片帝国所花费的金额令人震惊。2013年,清华紫光在国内掀起购物狂潮,斥资数十亿美元购买了两家中国最成功的无晶圆芯片设计公司展讯通信和锐迪科微电子,后者为智能手机生产低端芯片。赵宣布合并将产生“在中国和国外产生了巨大的协同效应”,尽管近十年来几乎没有证据表明任何协同效应已经实现。

Even given his fortune, which was estimated at $2 billion, the sums Zhao spent building his chip empire were shocking. In 2013, Tsinghua Unigroup started its shopping spree at home, spending several billion dollars buying two of China’s most successful fabless chip design companies, Spreadtrum Communications and RDA Microelectronics, which made low-end chips for smartphones. Zhao declared the merger would produce “enormous synergies in China and abroad,” though nearly a decade on there’s little evidence any synergies have materialized.

一年后的 2014 年,赵与英特尔达成协议,将英特尔的无线调制解调器芯片与紫光集团的智能手机处理器结合起来。英特尔希望此次合作能促进其在中国智能手机市场的销售,而赵希望他的公司能够学习英特尔的芯片设计专业知识。他对紫光集团的目标持开放态度:半导体是中国的“国家优先,”他说。与英特尔合作将“加速技术发展,进一步增强中国半导体公司的竞争力和市场地位”。

A year later, in 2014, Zhao cut a deal with Intel to couple Intel’s wireless modem chips with Tsinghua Unigroup’s smartphone processors. Intel hoped the tie-up would boost its sales in China’s smartphone market, while Zhao wanted his companies to learn from Intel’s chip design expertise. He was open about Tsinghua Unigroup’s goals: semiconductors were China’s “national priority,” he said. Working with Intel would “accelerate the technology development and further strengthen the competitiveness and market position of Chinese semiconductor companies.”

赵与英特尔的合作背后有一些商业逻辑,但许多其他决定似乎并不是出于盈利的愿望。例如,清华紫光集团提出资助试图打入 NAND 存储芯片市场的中国公司 XMC(后来被 YMTC 收购)。该公司的首席执行官在一次公开活动中承认,他最初要求 150 亿美元建造一座新工厂,但被告知要取而代之的是 240 亿美元,“基于如果他们要认真成为世界领导者,那么他们需要与世界领导者的投资相匹配。” 即使是在中国西部与赵一起长大的牧羊人也会意识到他正在肆无忌惮地发放数十亿美元的支票。后来有消息称,除了半导体,紫光集团也在投资房地产和在线赌博,这不足为奇。

Zhao’s partnership with Intel had some business logic behind it, but many other decisions didn’t appear driven by a desire to make a profit. For example, Tsinghua Unigroup offered to fund XMC (later acquired by YMTC), a Chinese firm trying to break into the NAND memory chip market. The company’s CEO admitted at one public event that he initially asked for $15 billion to build a new fab but was told to take $24 billion instead, “on the basis that if they were going to be serious about being a world leader then they needed to match the world leaders’ investment.” Even the goatherders Zhao grew up alongside in western China would have recognized he was handing out multibillion-dollar checks with reckless abandon. When it later emerged that in addition to semiconductors, Tsinghua Unigroup was also investing in real estate and online gambling, it was barely a surprise.

与此同时,中国国家支持的“大基金”,宣布计划向紫光集团首期投资超过 10 亿美元。这为该公司的战略提供了政府批准的印记。赵将他的努力转向海外。仅仅拥有中国的无晶圆厂公司或吸引外国公司在中国投资是不够的。他要控制世界芯片产业的制高点。他聘请了几位领先的台湾半导体高管,包括台湾第二大晶圆代工厂联电的前首席执行官。2015年,赵本山亲自访问台湾,敦促台湾取消对中国在芯片设计和制造等领域投资的限制。他购买了台湾的Powertech Technology 25%的股份,该公司组装和测试半导体,这是台湾规则允许的交易。他与台湾其他几家大型芯片组装商进行股份和合资企业。

China’s state-backed “Big Fund,” meanwhile, announced plans to invest an initial tranche of over $1 billion in Tsinghua Unigroup. This provided a stamp of government approval for the company’s strategy. Zhao turned his efforts overseas. It wasn’t enough to own China’s fabless companies or attract foreign firms to invest in China. He wanted to control the commanding heights of the world’s chip industry. He hired several leading Taiwanese semiconductor executives, including the former CEO of UMC, Taiwan’s second biggest foundry. In 2015, Zhao visited Taiwan himself and pressed the island to lift its restrictions on Chinese investment in sectors like chip design and fabrication. He bought a 25 percent stake in Taiwan’s Powertech Technology, which assembles and tests semiconductors, a transaction that was allowed under Taiwan’s rules. He pursued stakes and joint ventures with several of Taiwan’s other large chip assemblers.

然而,赵的真正兴趣在于购买岛上的皇冠上的明珠——美国以外领先的芯片设计公司联发科和全球几乎所有无晶圆厂芯片公司都依赖的代工厂台积电。他提出了购买台积电 25% 股份的想法,并主张将联发科与紫光集团的芯片设计业务合并。根据台湾现行的外商投资规则,这两项交易都不合法,但当赵从台湾回来时,他在北京的一个公开会议上登台,并建议如果台北不改变这些限制,中国应该禁止进口台湾芯片。

However, Zhao’s real interest was in buying the island’s crown jewels—MediaTek, the leading chip designer outside the U.S., and TSMC, the foundry on which almost all the world’s fabless chip firms rely. He floated the idea of buying a 25 percent stake in TSMC and advocated merging MediaTek with Tsinghua Unigroup’s chip design businesses. Neither transaction was legal under Taiwan’s existing foreign investment rules, but when Zhao returned from Taiwan he took the stage at a public conference in Beijing and suggested China should ban imports of Taiwanese chips if Taipei didn’t change these restrictions.

这场施压运动使台积电和联发科陷入了困境。两家公司都严重依赖中国市场。台积电生产的大部分芯片都在中国各地的车间组装成电子产品。将台湾的科技皇冠上的宝石卖给大陆有国家支持的投资者的想法毫无意义。该岛最终将依赖北京。除了废除军队或欢迎人民解放军占领外,很难想象有什么步骤会更有利于破坏台湾的自治。

This pressure campaign put TSMC and MediaTek in a bind. Both companies were crucially reliant on the Chinese market. Most of the chips TSMC produced were assembled into electronics goods in workshops across China. The idea of selling Taiwan’s technological crown jewels to a state-backed investor on the mainland made little sense. The island would end up dependent on Beijing. Besides abolishing its military or welcoming occupation by the People’s Liberation Army, it was hard to think of a step that would do more to undermine Taiwan’s autonomy.

台积电和联发科均发表声明,含糊表达对中国投资的开放态度。莫里斯·张说他唯一的规定是“如果价格合适,是否对股东有利”——对于一项可能会破坏台湾经济独立的交易,人们几乎不会想到会有这样的反应。但张也警告说,如果中国投资者可以任命台湾公司董事会成员,“保护知识分子不会那么容易财产。” 联发科表示支持“携手提升中台企业在全球芯片产业的地位和竞争力”——但前提是台湾政府允许。然而,在台北,政府似乎在摇摆不定。台湾经济部长邓小平建议放宽台湾对中国在芯片领域投资的限制。在中国的压力下,他暗示中国对台湾芯片行业的更大控制是不可避免的。“你不能逃避这个问题,”邓告诉记者。但在台湾有争议的总统选举中,政府推迟了任何政策变化。

Both TSMC and MediaTek issued statements vaguely expressing openness to Chinese investment. Morris Chang said his only stipulations were “if the price is right and if it is beneficial to shareholders”—hardly the response one would expect about a deal that threatened to undermine Taiwan’s economic independence. But Chang also warned that if Chinese investors could appoint members to Taiwanese companies’ boards of directors, “it will not be that easy to protect intellectual property.” MediaTek said it was supportive of efforts “to join hands and raise the status and competitiveness of the Chinese and Taiwanese enterprises in the global chip industry”—but only if the Taiwanese government allowed. In Taipei, the government seemed to be wobbling, however. John Deng, the island’s economy minister, suggested relaxing Taiwan’s restrictions on Chinese investment in the chip sector. Amid Chinese pressure, he signaled that greater Chinese control of Taiwan’s chip sector was inevitable. “You cannot escape from this issue,” Deng told journalists. But amid a contentious presidential election in Taiwan, the government delayed any policy changes.

很快,赵将目光投向了美国的半导体产业。2015年7月,清华紫光提出了以 230 亿美元收购美国存储芯片生产商美光的想法,这将是中国在任何行业中对美国公司的最大一笔收购。与台湾科技巨头及其经济专家的情况不同,清华收购美光的努力遭到坚决拒绝。美光表示,鉴于该交易并不现实美国政府的安全担忧。不久之后,2015年9月,清华紫光再次尝试,延长 37 亿美元收购另一家制造 NAND 存储芯片的美国公司 15% 股份的报价。评估外国投资的美国政府机构 CFIUS 以安全为由拒绝了这一做法。

Soon Zhao set his sights on America’s semiconductor industry. In July 2015, Tsinghua Unigroup floated the idea of buying Micron, the American memory chip producer, for $23 billion, which would have been the largest ever Chinese purchase of a U.S. company in any industry. Unlike in the case of Taiwan’s tech titans and its economic technocrats, Tsinghua’s efforts to purchase Micron were firmly rebuffed. Micron said it didn’t think the transaction was realistic given the U.S. government’s security concerns. Soon after, in September 2015, Tsinghua Unigroup tried again, extending a $3.7 billion offer for a 15 percent stake in another U.S. company that made NAND memory chips. CFIUS, the U.S. government body that assesses foreign investment, rejected this on security grounds.

然后,在 2016 年春季,清华悄悄收购了另一家美国芯片公司莱迪思半导体 6% 的股份。“这纯粹是一种金融投资,”赵告诉华尔街日报。“我们根本没有任何试图收购莱迪思的意图。” 几乎没有投资公布几周后,紫光集团开始出售其在莱迪思的股份。此后不久,Lattice 收到了来自加州一家名为 Canyon Bridge 的投资公司的收购要约,路透社的记者透露,由中国政府谨慎资助。美国政府坚决拒绝这笔交易。

Then, in spring 2016, Tsinghua quietly bought 6 percent of the shares in Lattice Semiconductor, another U.S. chip firm. “This is purely a financial investment,” Zhao told the Wall Street Journal. “We don’t have any intention at all to try to acquire Lattice.” Scarcely weeks after the investment was publicized, Tsinghua Unigroup began to sell its shares in Lattice. Shortly thereafter, Lattice received a buyout offer from a California-based investment firm called Canyon Bridge, which journalists from Reuters revealed had been discreetly funded by the Chinese government. The U.S. government firmly rejected the deal.

同一个投资基金同时收购了陷入财务困境的英国芯片设计公司 Imagination。该交易是精心设计以排除 Imagination 在美国的资产这样华盛顿也没有阻止它。英国监管机构挥手批准了这笔交易,但三年后,当新东家试图重组董事会,由中国政府投资基金任命的官员组成。

The same investment fund simultaneously bought Imagination, a UK-based chip designer in financial distress. The transaction was carefully structured to exclude Imagination’s U.S. assets so that Washington didn’t block it, too. British regulators waved the deal through, only to find themselves regretting the decision when, three years later, the new owners tried to restructure the board of directors with officials appointed by a Chinese government investment fund.

问题不仅仅是中国政府相关基金正在收购外国芯片公司。他们这样做的方式违反了有关市场操纵和内幕交易。例如,当 Canyon Bridge 打算收购莱迪思半导体时,Canyon Bridge 的一位联合创始人向北京的一位同事透露,通过微信和在北京星巴克的会议上传递了有关交易的细节。他的同事根据这些知识买了股票;Canyon Bridge 高管被判内幕交易罪名成立。

The problem wasn’t simply that Chinese government-linked funds were buying up foreign chip firms. They were doing so in ways that violated laws about market manipulation and insider trading. While Canyon Bridge was maneuvering to purchase Lattice Semiconductor, for example, one of Canyon Bridge’s cofounders tipped off a colleague in Beijing, passing along details about the transaction via WeChat and at meetings in a Starbucks in Beijing. His colleague bought stock based on this knowledge; the Canyon Bridge executive was convicted of insider trading.

就他而言,赵认为自己只是一个坚定的企业家。“美国和中国大公司之间的合并是一定会发生,”他宣称。“应该从商业角度看待它们,而不是在民族主义或政治背景下对待它们。” 但紫光集团的活动从商业逻辑的角度是无法理解的。有太多的中国国有和国家资助的“私募股权”公司环绕着世界半导体公司,无法将其描述为政府主导的抢占外国芯片公司的努力。“发动突击,”习近平曾要求。赵、清华紫光和其他政府支持的“投资”机构只是按照这些公开宣布的指示行事。在这场疯狂的交易中,清华紫光集团在 2017 年宣布,获得了新的“投资”:来自国家开发银行的约 150 亿美元和来自集成电路产业投资基金的 70 亿美元——两者均由中国政府拥有和控制。

For his part, Zhao saw himself as simply a committed entrepreneur. “Mergers between big U.S. and Chinese companies are bound to happen,” he declared. “They should be viewed from a business perspective instead of being treated under nationalist or political contexts.” But Tsinghua Unigroup’s activities were impossible to comprehend from the perspective of business logic. There were too many Chinese state-owned and state-financed “private equity” firms circling the world’s semiconductor companies to describe this as anything other than a government-led effort to seize foreign chip firms. “Call forth the assault,” Xi Jinping had demanded. Zhao, Tsinghua Unigroup, and other government-backed “investment” vehicles were simply following these publicly announced instructions. Amid this frenzied dealmaking, Tsinghua Unigroup announced in 2017 that it had received new “investment”: around $15 billion from the China Development Bank and $7 billion from the Integrated Circuit Industry Investment Fund—both owned and controlled by the Chinese state.

第 46 章 华为的崛起

CHAPTER 46 The Rise of Huawei

W任正非在他创立的中国科技公司华为总部接受媒体采访时,剪裁利落的夹克和休闲裤、解开的衣领、活泼的笑容让他看起来就像任何硅谷高管。在某些方面他是。他的公司的电信设备——手机信号塔上的无线电,用于在智能手机之间传输电话、图片和电子邮件——构成了世界移动互联网的支柱。与此同时,直到最近,华为的智能手机部门还是全球最大的部门之一,在售出的手机数量上可与苹果和三星相媲美。该公司还提供其他类型的技术基础设施,从海底光缆到云计算。在许多国家,不使用华为的某些设备就无法使用手机——就像使用不带 Microsoft 产品的 PC 或不使用 Google 上网(中国以外)一样困难。然而,华为与世界上其他大型科技公司有一个主要区别:它与美国国家安全国家长达两个十年的斗争。

When Ren Zhengfei gives media interviews in the headquarters of Huawei, the Chinese technology company he founded, his crisply tailored jacket and slacks, unbuttoned collar, and vivacious smile make him seem just like any Silicon Valley executive. In some ways he is. His company’s telecom equipment—the radios on cell towers that transmit calls, pictures, and emails to and from smartphones—forms the backbone of the world’s mobile internet. Huawei’s smartphone unit, meanwhile, was until recently one of the world’s largest, rivaling Apple and Samsung in numbers of phones sold. The company provides other types of tech infrastructure, too, from undersea fiber-optic cables to cloud computing. In many countries it’s impossible to use a phone without using some of Huawei’s equipment—as difficult as it is to use a PC without Microsoft products or to surf the internet (outside of China) without Google. However, Huawei is different from the world’s other big tech companies in one major way: its two-decade-long struggle with America’s national security state.

阅读有关华为在中国政府间谍活动中的作用的美国报纸头条新闻,很容易得出结论,该公司是中国安全机构的附属机构。这华为与中国政府之间的关系有据可查,但很少解释该公司如何建立全球业务。要了解该公司的扩张,将华为的发展轨迹与另一家专注于科技的企业集团韩国三星进行比较会更有帮助。任正非比三星的李秉哲晚一代出生,但两位大亨的运营模式相似。Lee 将三星从一家干鱼贸易商变成了一家科技公司,依靠三种策略生产出一些世界上最先进的处理器和内存芯片。一是积极培育政治关系,争取有利监管和廉价资本。其次,找出西方和日本首创的产品,并学习以同等质量和更低的成本制造它们。第三,坚持全球化,不仅要寻找新客户,还要通过与世界上最好的公司竞争来学习。执行这些战略使三星成为世界上最大的公司之一,

Reading American newspaper headlines about Huawei’s role in Chinese government spying, it would be easy to conclude that the company emerged as an appendage of China’s security agencies. The ties between Huawei and the Chinese state are well documented but explain little about how the company built a globe-spanning business. To understand the company’s expansion, it’s more helpful to compare Huawei’s trajectory to a different tech-focused conglomerate, South Korea’s Samsung. Ren was born a generation after Samsung’s Lee Byung-Chul, but the two moguls have a similar operating model. Lee built Samsung from a trader of dried fish into a tech company churning out some of the world’s most advanced processor and memory chips by relying on three strategies. First, assiduously cultivate political relationships to garner favorable regulation and cheap capital. Second, identify products pioneered in the West and Japan and learn to build them at equivalent quality and lower cost. Third, globalize relentlessly, not only to seek new customers but also to learn by competing with the world’s best companies. Executing these strategies made Samsung one of the world’s biggest companies, achieving revenues equivalent to 10 percent of South Korea’s entire GDP.

一家中国公司能否执行一套类似的战略?中国的大多数科技公司都尝试了一种不同的方法,但对全球的关注度较低。尽管中国的出口实力如此强大,但中国的互联网公司几乎所有的钱都是在中国国内市场上赚钱的,在那里它们受到监管和审查的保护。腾讯、阿里巴巴、拼多多和美团如果不是因为它们在国内市场的主导地位,就会成为小鱼。当中国科技公司走出国门时,他们常常难以竞争。

Could a Chinese firm execute a similar set of strategies? Most of China’s tech firms tried a different approach with a less global focus. For all the country’s export prowess, China’s internet firms make almost all their money inside of China’s domestic market, where they’re protected by regulation and censorship. Tencent, Alibaba, Pinduoduo, and Meituan would be minnows were it not for their home market dominance. When Chinese tech firms have gone abroad, they’ve often struggled to compete.

相比之下,华为从一开始就接受外国竞争。任正非的商业模式与阿里巴巴或腾讯有着根本的不同。他采用国外首创的概念,以较低的成本生产优质版本,并销往世界,从国际竞争对手那里抢占国际市场份额。这种商业模式使三星的创始人变得富有,并将公司置于世界科技生态系统的中心。直到最近,华为似乎也走上了同样的道路。

By contrast, Huawei has embraced foreign competition from its earliest days. Ren Zhengfei’s business model has been fundamentally different from Alibaba’s or Tencent’s. He’s taken concepts pioneered abroad, produced quality versions at lower cost, and sold them to the world, grabbing international market share from international rivals. This business model made Samsung’s founders rich and put the company at the center of the world’s tech ecosystem. Until very recently, Huawei seemed to be on the same path.

公司的国际化定位从1987年成立就可见一斑。任先生在中国南方贵州省农村的一个高中教师家庭长大。他接受过工程师培训在中国军队服役之前,他曾在四川省会重庆工作,他说他在那里的一家工厂工作服装用合成纤维。据报道,在离开军队后(一些怀疑论者怀疑当时的情况,如果他真的与军方完全断绝关系),他搬到了深圳,当时是一个与香港接壤的小镇。当时,香港仍然由英国人统治,这是一个繁荣的小前哨,位于原本贫穷的华南沿岸。大约十年前,中国领导人开始实施经济改革,尝试让个人组建私营公司,以此作为刺激经济增长的手段。深圳是几个被选为“经济特区”的城市之一,这些城市取消了限制性法律,鼓励外商投资。

The company’s international orientation was visible from its founding in 1987. Ren had grown up in a family of high school teachers in rural Guizhou Province in southern China. He’d trained as an engineer in Sichuan’s capital of Chongqing before serving in the Chinese army, where he says he worked in a factory producing synthetic fiber for garments. After reportedly leaving the army (some skeptics wonder about the circumstances, and if he actually did cut ties with the military completely), he moved to Shenzhen, then a small town just across the border from Hong Kong. At the time, Hong Kong was still ruled by the British, a small outpost of prosperity along the otherwise impoverished South China coast. China’s leaders had begun implementing economic reforms about a decade earlier, experimenting with letting individuals form private companies as a means of spurring economic growth. Shenzhen was one of several cities selected as a “special economic zone,” where restrictive laws were canceled and foreign investment was encouraged. The city boomed as Hong Kong money flowed in and as China’s would-be entrepreneurs flocked to the city in search of freedom from regulation.

任先生看到了进口电信交换机的机会,这种设备将一个呼叫者连接到另一个呼叫者。凭借 5,000 美元的启动资金,他开始从香港进口这种装备。当他在边境的伙伴意识到他通过转售他们的设备赚了很多钱时,他们切断了他的联系,所以任决定自己制造设备。到 1990 年代初,华为有数百人从事研发工作,主要集中在楼宇交换设备。从那时起,电信基础设施已与数字基础设施合并。传输呼叫的相同蜂窝塔也发送其他类型的数据。因此,华为的设备现在在传输世界数据方面发挥着重要的——在许多国家也是至关重要的——作用。今天,它与芬兰的诺基亚和瑞典的爱立信并驾齐驱,是世界三大手机信号塔设备供应商之一。

Ren saw an opportunity to import telecom switches, the equipment that connects one caller to another. With $5,000 in startup capital, he began importing this gear from Hong Kong. When his partners across the border realized he was making good money by reselling their equipment, they cut him off, so Ren decided to build his own equipment. By the early 1990s, Huawei had several hundred people working in R&D, largely focused on building switching equipment. Since those days, the telecom infrastructure has merged with digital infrastructure. The same cell towers that transmit calls also send other types of data. So Huawei’s equipment now plays an important—and in many countries, crucial—role in transmitting the world’s data. Today it is one of the world’s three biggest providers of equipment on cell towers, alongside Finland’s Nokia and Sweden’s Ericsson.

华为的批评者经常声称其成功建立在知识产权被盗的基础上,尽管这只是部分正确。该公司承认了一些先前的知识产权侵权行为,并被指控更多。例如,2003 年,华为承认其一台路由器中 2% 的代码是直接从思科复制,美国竞争对手。与此同时,加拿大报纸报道称该国的间谍机构认为,在 2000 年代,有中国政府支持的针对加拿大电信巨头北电的黑客和间谍活动,据称这使华为受益。

Huawei’s critics often allege that its success rests on a foundation of stolen intellectual property, though this is only partly true. The company has admitted to some prior intellectual property violations and has been accused of far more. In 2003, for example, Huawei acknowledged that 2 percent of the code in one of its routers was copied directly from Cisco, an American competitor. Canadian newspapers, meanwhile, have reported that the country’s spy agencies believe there was a Chinese-government-backed campaign of hacking and espionage against Canadian telecom giant Nortel in the 2000s, which allegedly benefitted Huawei.

盗窃知识产权可能使公司受益,但不能解释它的成功。再多的知识产权或商业秘密也不足以建立像华为这样大的企业。该公司开发了高效的制造工艺,降低了成本并制造了客户认为高质量的产品。与此同时,华为在研发上的投入处于世界领先地位。该公司在研发上的投入是其他中国科技公司的数倍。它大致每年 150 亿美元的研发预算只有少数几家公司拥有,包括像谷歌和亚马逊这样的科技公司,像默克这样的制药公司,以及像戴姆勒或大众这样的汽车制造商。即使衡量华为在知识产权盗窃方面的记录,该公司数十亿美元的研发支出表明,与苏联泽列诺格勒或许多其他试图打入芯片行业的中国公司的“抄袭”心态有着根本不同。便宜的。

Theft of intellectual property may well have benefitted the company, but it can’t explain its success. No quantity of intellectual property or trade secrets is enough to build a business as big as Huawei. The company has developed efficient manufacturing processes that have driven down costs and built products that customers see as high-quality. Huawei’s spending on R&D, meanwhile, is world leading. The company spends several times more on R&D than other Chinese tech firms. Its roughly $15 billion annual R&D budget is paralleled by only a handful of firms, including tech companies like Google and Amazon, pharmaceutical companies like Merck, and carmakers like Daimler or Volkswagen. Even when weighing Huawei’s track record of intellectual property theft, the company’s multibillion-dollar R&D spending suggests a fundamentally different ethos than the “copy it” mentality of Soviet Zelenograd, or the many other Chinese firms that have tried to break into the chip industry on the cheap.

华为高管表示,他们投资研发是因为他们向硅谷学习。据报道,任正非带了一群华为高管1997 年访问美国,参观了惠普、IBM 和贝尔实验室等公司。他们离开时坚信研发的重要性,以及有效的管理流程的重要性。从 1999 年开始,华为聘请了 IBM 的咨询部门来教它像世界级公司一样运作。一位前 IBM 顾问表示,华为 1999 年在咨询费上花费了 5000 万美元,而当时其全部收入还不到 10 亿美元。有一次,它雇佣了 100 名 IBM 员工来重做业务流程。“他们并没有被工程任务吓倒,”这位前顾问报告说,但“他们觉得在经济知识和商业知识方面落后了一百年。” 多亏了IBM和其他西方顾问,华为学会了管理供应链、预测客户需求、开发一流的营销和全球销售产品。

Huawei executives say they invest in R&D because they’ve learned from Silicon Valley. Ren reportedly brought a group of Huawei executives to tour the U.S. in 1997, visiting companies like HP, IBM, and Bell Labs. They left convinced of the importance not only of R&D, but also of effective management processes. Starting in 1999, Huawei hired IBM’s consulting arm to teach it to operate like a world-class company. One former IBM consultant said Huawei spent $50 million in 1999 on consulting fees, at a time when its entire revenue was less than a billion dollars. At one point it employed one hundred IBM staff to redo business processes. “They weren’t too daunted by the engineering tasks,” this former consultant reported, but “they felt they were a hundred years behind when it came to economic knowledge and business knowledge.” Thanks to IBM and other Western consultants, Huawei learned to manage its supply chain, anticipate customer demand, develop top-class marketing, and sell products worldwide.

华为将此与该公司称为“狼文化”的军国主义精神相结合。公司一间研究室墙上的书法写着“牺牲是军人的最高事业。胜利是士兵最大的贡献,”据《纽约时报》报道。然而,在芯片行业的背景下,任正非的军国主义并不是那么独特。安迪·格鲁夫(Andy Grove)写了一本关于偏执狂的好处的畅销书。与此同时,Morris Chang 说他会研究了二战中最血腥的斯大林格勒战役,以获取有关商业的课程。

Huawei coupled this with a militaristic ethos that the company celebrates as “wolf-culture.” Calligraphy on the wall of one of the company’s research lab reads “Sacrifice is a soldier’s highest cause. Victory is a soldier’s greatest contribution,” according to a New York Times report. In the context of the chip industry, though, Ren Zhengfei’s militarism wasn’t that unique. Andy Grove wrote a bestseller about the benefits of paranoia. Morris Chang, meanwhile, said that he’d studied Stalingrad, the bloodiest battle of World War II, for lessons about business.

除了西方咨询公司,华为还得到了另一个强大机构的帮助:中国政府。在发展的不同阶段,华为都受益于深圳地方政府、国有银行和北京中央政府的支持。《华尔街日报》对中国政府提供的补贴总额的审查达到了750 亿美元,以补贴土地、国家支持的信贷和税收减免的形式,其规模远高于大多数西方公司从其政府那里获得的收益,尽管提供给华为的收益可能与其他东亚政府提供的收益差别不大优先公司。

In addition to Western consulting firms, Huawei had help from another powerful institution: China’s government. At different points in its development, Huawei has benefitted from support from the local government in Shenzhen, from state-owned banks, and from the central government in Beijing. A Wall Street Journal review of total subsidies provided by the Chinese government reached a figure of $75 billion, in the form of subsidized land, state-backed credit, and tax deductions at a scale far above what most Western companies get from their governments, though the benefits provided to Huawei might not be too different from what other East Asian governments provide to priority companies.

国家对一家表面上是私营公司的支持规模已经引发了危险信号,尤其是在美国。中国领导人当然一直支持该公司的全球扩张。即使在 1990 年代中期,当华为还是一家小公司时,吴邦国副总理等中国高级官员参观了该公司并承诺给予支持。吴副总理还随任正非出国,帮助华为在非洲销售电信设备。然而,鉴于中国对国际贸易采取的重商主义态度以及公私财产之间的模糊界限,很难区分这是对华为的特殊支持,还是仅仅是标准操作程序。

The scale of state support for an ostensibly private firm has raised red flags, especially in the United States. China’s leaders have certainly been supportive of the company’s global expansion. Even in the mid-1990s, when Huawei was still a small company, top Chinese officials like Vice Premier Wu Bangguo visited the company and promised to support it. Vice Premier Wu also traveled abroad with Ren Zhengfei to help Huawei sell telecom equipment in Africa. Yet it’s hard to distinguish whether this amounted to special support for Huawei or was simply standard operating procedure given China’s mercantilist approach to international trade and the fuzzy boundaries between public and private property.

任正非从中国人民解放军到华为的过渡缺乏明确性仍然令人费解。该公司复杂而不透明的股权结构也引发了合理的质疑。华为高管胡肯在美国国会调查中称任正非加入中国共产党就像“一些美国商人民主党或共和党,”在美国分析师看来,这听起来像是故意混淆共产党在公司治理中的作用。然而,华为是中国政府专门建造的论点从来没有强有力的证据支持。

The lack of clarity about Ren’s transition from the People’s Liberation Army to Huawei remains puzzling. The company’s complex and opaque ownership structure has also provoked reasonable questions. Huawei executive Ken Hu’s argument to a U.S. congressional inquiry that Ren Zhengfei’s membership in the Chinese Communist Party was just like how “some American businessmen are Democrat or Republican,” sounded to U.S. analysts like willful obfuscation of the Communist Party’s role in the company’s governance. Nevertheless, the thesis that Huawei was purpose built by the Chinese state has never had strong evidence behind it.

然而,华为的崛起符合中国政府的利益,因为该公司抢占了市场份额,并将其设备嵌入了全球电信网络。多年来,不顾美国间谍机构的警告,华为在全球迅速蔓延。随着它的发展,出售电信设备的现有西方公司被迫合并或被赶出市场。加拿大的北电破产了。在 AT&T 分拆后继承贝尔实验室的公司阿尔卡特朗讯将其业务出售给了芬兰的诺基亚。

Huawei’s rise has, however, worked in the interests of the Chinese state, as the company grabbed market share and embedded its equipment in the world’s telecom networks. For many years, despite the warning of America’s spy agencies, Huawei spread rapidly across the world. As it grew, incumbent Western firms selling telecom equipment were forced to merge or pushed out of the market. Canada’s Nortel went bankrupt. Alcatel-Lucent, the company that inherited Bell Labs after AT&T was broken up, sold its operations to Finland’s Nokia.

华为的野心只会越来越大。在提供了使电话成为可能的基础设施之后,它也开始销售电话。很快,它的智能手机就跻身世界畅销书之列。到 2019 年,按销量衡量,该公司仅落后于三星。与三星或苹果相比,华为每部手机的收入仍然要少得多,后者的营销和生态系统可以收取更高的价格。然而,华为进入智能手机市场并迅速占据领先地位的能力让苹果和三星备受瞩目。

Huawei’s ambitions only grew. Having provided the infrastructure that makes phone calls possible, it started selling phones, too. Soon its smartphones were among the world’s best sellers. By 2019 the company lagged only Samsung measured by number of units sold. Huawei still made substantially less money per phone than either Samsung or Apple, the latter of which had the marketing and the ecosystem to charge vastly higher prices. However, Huawei’s ability to enter the smartphone market and quickly seize a leading position put Apple and Samsung on notice.

此外,华为在设计自己手机中的一些关键芯片方面取得了进展。公司内部人士表示,该公司的芯片设计雄心在 2011 年 3 月加速,当时日本东海岸的地震引发了席卷日本的海啸。全世界的注意力都集中在被洪水破坏的福岛第一核反应堆上,但在华为内部,高管们担心该公司的供应链会受到威胁。与所有主要的电子产品生产商一样,华为依赖日本供应商为其电信设备和智能手机提供关键组件,并担心这场灾难可能会造成巨大的延误。最终,华为走运了。很少有零部件供应商长期停产。然而,该公司要求其顾问确定其供应链风险。他们报告说,该公司有两个关键漏洞:访问谷歌的 Android 操作系统、所有非苹果智能手机运行的核心软件,以及每部智能手机所需的半导体供应。

Moreover, Huawei was making progress designing some of the critical chips in its own phones. Company insiders say the firm’s chip design ambitions accelerated in March 2011, when an earthquake off Japan’s east coast caused a tsunami that slammed into the country. The world’s attention focused on the Fukushima Daiichi nuclear reactor that was damaged by the flooding, but inside Huawei, executives worried about the threat to the company’s supply chain. Like every major producer of electronics, Huawei relied on Japanese providers for crucial components in their telecom gear and smartphones and feared the disaster might cause immense delays. In the end, Huawei got lucky. Few of its component suppliers saw production knocked out for long. However, the company asked its consultants to determine its supply chain risk. They reported that the company had two key vulnerabilities: access to Google’s Android operating system, the core software on which all non-Apple smartphones run, and the supply of the semiconductors that every smartphone requires.

该公司确定了其产品所需的 250 种最重要的半导体并开始尽可能多地在内部设计。这些芯片主要与建造电信基站业务有关,但也包括公司智能手机的应用处理器,这些半导体极其复杂,需要最先进的芯片制造技术。与苹果和大多数其他领先的芯片公司一样,华为选择外包这些芯片的制造,因为它需要使用最多几家公司可以提供的制造工艺。台湾的台积电自然是转向的地方。

The company identified the 250 most important semiconductors that its products required and began designing as many as possible in-house. These chips were largely related to the business of building telecom base stations but also included the application processors for the company’s smartphones, semiconductors that were monstrously complex and required the most advanced chipmaking technology. Like Apple and most other leading chip firms, Huawei chose to outsource fabrication of these chips, because it needed to use manufacturing processes that, at most, a couple companies could provide. Taiwan’s TSMC was the natural place to turn.

到 2010 年代末,华为的海思部门正在设计一些世界上最复杂的智能手机芯片,并已成为台积电的第二大客户。华为的手机仍然需要其他公司的芯片,比如内存芯片或各种类型的信号处理器。但掌握手机处理器的生产是一项了不起的壮举。美国对世界上最赚钱的芯片设计业务的近乎垄断地位受到威胁。这更加证明了华为成功地复制了韩国三星或日本索尼几十年前所做的事情:学习生产先进技术、赢得全球市场、投资研发以及挑战美国的技术领导者。此外,随着下一代电信基础设施 5G 的推出,华为似乎在一个无处不在的计算新时代处于独特的有利地位。

By the end of the 2010s, Huawei’s HiSilicon unit was designing some of the world’s most complex chips for smartphones and had become TSMC’s second-largest customer. Huawei’s phones still required chips from other companies, too, like memory chips or various types of signal processors. But mastering the production of cell phone processors was an impressive feat. America’s near monopoly on the world’s most profitable chip design businesses was under threat. This was more evidence that Huawei was successfully replicating what South Korea’s Samsung or Japan’s Sony had done decades earlier: learning to produce advanced technology, winning global markets, investing in R&D, and challenging America’s tech leaders. Moreover, Huawei seemed uniquely well placed for a new era of ubiquitous computing that would accompany the rollout of the next generation of telecom infrastructure: 5G.

第 47 章 5G 未来

CHAPTER 47 The 5G Future

当任正非开始从香港进口电话交换机时,除了将一部电话连接到另一部电话外,网络设备无能为力。在电话的早期,切换是手动完成的,一排排女性坐在一排插头的前面,根据打电话的人以不同的组合连接它们。到 1980 年代,人类已被电子开关所取代,而电子开关通常依赖于半导体设备。即便如此,还是花了转换一个壁橱大小的设备来管理一座建筑物的电话线。今天,电信供应商比以往任何时候都更加依赖硅,但一个壁橱般的设备可以处理呼叫、文本和视频,现在通常通过无线电网络而不是固定电话发送。

When Ren Zhengfei started importing telephone switches from Hong Kong, network gear couldn’t do much beyond connecting one phone to another. In the early days of telephones, switching had been done by hand, with rows of women seated in front of a wall of plugs, connecting them in different combinations depending on who was calling. By the 1980s, humans had been replaced by electronic switches, which often relied on semiconductor devices. Even still, it took switching gear the size of a closet to manage a single building’s worth of telephone lines. Today, telecom providers are more reliant than ever on silicon, but a closet’s worth of gear can process calls, texts, and video, now often sent via radio networks rather than landlines.

华为已经掌握了通过蜂窝网络发送呼叫和数据的最新一代设备,称为 5G。然而,5G 并不是真正的手机——它是关于计算的未来,因此,它是关于半导体的。5G中的“G”代表世代。我们已经循环了四代移动网络标准,每一代都需要手机和手机信号塔上的新硬件。正如摩尔定律允许我们在芯片上封装更多晶体管一样,通过无线电波往返手机的 1 和 0 的数量也在稳步增加。2G手机可以发图文;3G手机打开的网站;4G 使得几乎可以从任何地方流式传输视频。5G 将提供类似的飞跃。

Huawei has mastered the latest generation of equipment to send calls and data via cell networks, called 5G. Yet 5G isn’t really about phones—it’s about the future of computing, and therefore, it’s about semiconductors. The “G” in 5G stands for generation. We’ve already cycled through four generations of mobile networking standards, each of which required new hardware on phones and in cell towers. Just as Moore’s Law has let us pack more transistors onto chips, there’s been a steady increase in the number of 1s and 0s flying to and from cell phones via radio waves. 2G phones could send picture texts; 3G phones opened websites; and 4G made it possible to stream video from almost anywhere. 5G will provide a similar leap forward.

今天的大多数人都认为他们的智能手机是理所当然的,但正是由于功能越来越强大的半导体,我们不再惊叹于图片文本,而是对视频流的瞬间延迟感到沮丧。管理手机与蜂窝网络连接的调制解调器芯片可以通过手机天线在无线电波中发送更多的 1 和 0。

Most people today take their smartphone for granted, but it’s only thanks to ever more powerful semiconductors that we no longer marvel at picture texts and are instead frustrated with split-second delays in video streaming. The modem chips that manage a phone’s connection with cell networks make it possible to send many more 1s and 0s in the radio waves via a phone’s antenna.

隐藏在蜂窝网络内和蜂窝塔顶部的芯片也发生了类似的变化。通过空中发送 1 和 0,同时最大限度地减少掉线或视频流延迟是非常复杂的。无线电波频谱相关部分的可用空间量是有限的。无线电波频率只有这么多,其中许多都不是发送大量数据或长距离传输的最佳选择。因此,电信公司依靠半导体将更多数据打包到现有频谱空间中。“频谱比硅贵得多,”Analog Devices 的芯片专家 Dave Robertson 解释说,该公司专门研究管理无线电传输的半导体。因此,半导体对于无线发送更多数据的能力至关重要。更精确,同时使用更少的功率。

There’s been a comparable change in the chips hidden inside a cell network and atop cell towers. Sending 1s and 0s through the air while minimizing dropped calls or delays to video streaming is staggeringly complicated. The amount of space available in the relevant part of the radio-wave spectrum is limited. There are only so many radio-wave frequencies, many of which aren’t optimal for sending lots of data or transmitting over long distances. Telecom firms have therefore relied on semiconductors to pack ever more data into existing spectrum space. “Spectrum is far more expensive than silicon,” explains Dave Robertson, a chip expert at Analog Devices, which specializes in semiconductors that manage radio transmission. Semiconductors have therefore been fundamental to the ability to send more data wirelessly. Chip designers like Qualcomm found new ways to optimize transmission of data via the radio spectrum, and chipmakers like Analog Devices have made semiconductors called radio frequency transceivers that can send and receive radio waves with more precision while using less power.

下一代网络技术 5G 将使更多数据的无线传输成为可能。在某种程度上,这将通过更复杂的频谱空间共享方法实现,这需要更复杂的算法和手机和手机信号塔的更多计算能力,以便即使是无线频谱中最小的空闲空间也可以插入 1 和 0。部分原因是,5G 网络将通过使用新的、空的无线电频谱发送更多数据,而该频谱以前被认为是不切实际的。先进的半导体不仅可以将更多的 1 和 0 装入给定频率的无线电波中,而且还可以将无线电波发送得更远,并以前所未有的准确度瞄准它们。蜂窝网络将识别手机的位置,并使用一种称为波束成形的技术直接向手机发送无线电波。典型的无线电波,例如向您的汽车收音机发送音乐的无线电波,会向各个方向发送信号,因为它不知道您的汽车在哪里。这会浪费电力并产生更多的波和更多的干扰。通过波束成形,蜂窝塔可以识别设备的位置并仅在该方向发送它需要的信号。结果:每个人的干扰更少,信号更强。

The next generation of network technology, 5G, will make possible the wireless transmission of even more data. Partly, this will be via even more intricate methods of sharing spectrum space, which require more complex algorithms and more computing power on phones and in cell towers so that 1s and 0s can be slotted in even the tiniest free space in the wireless spectrum. Partly, 5G networks will send more data by using a new, empty radio frequency spectrum that was previously considered impractical to fill. Advanced semiconductors make it possible not only to pack more 1s and 0s into a given frequency of radio waves, but also to send radio waves farther and target them with unprecedented accuracy. Cell networks will identify a phone’s location and send radio waves directly toward a phone, using a technique called beamforming. A typical radio wave, like one that sends music to your car radio, sends signals out in every direction because it doesn’t know where your car is. This wastes power and creates more waves and more interference. With beamforming, a cell tower identifies a device’s location and sends the signal it needs only in that direction. Result: less interference and stronger signals for everyone.

能够承载更多数据的更快网络不仅能让现有手机运行得更快——它们还将改变我们对移动计算的看法。在 1G 网络时代,手机对大多数人来说太贵了。有了 2G 网络,我们开始假设手机可以发送短信和语音。今天,我们希望手机和平板电脑几乎具备 PC 的所有功能。随着通过蜂窝网络发送更多数据成为可能,我们将越来越多的设备连接到蜂窝网络。我们拥有的设备越多,它们产生的数据就越多,这需要更多的处理能力才能理解。

Faster networks capable of carrying more data won’t simply let existing phones run faster—they’ll change the way we think about mobile computing. In the age of 1G networks, cell phones were too expensive for most people to own. With 2G networks, we came to assume that phones could send text messages as well as voice. Today, we expect phones and tablets to have almost all the features of PCs. As it becomes possible to send even more data over cell networks, we’ll connect ever more devices to the cell network. The more devices we have, the more data they’ll produce, which will require more processing power to make sense of.

将更多设备连接到蜂窝网络并从中收集数据的承诺听起来可能不是革命性的。您可能认为 5G 网络无法冲泡出更好的咖啡,但用不了多久,您的咖啡机就会收集和处理其生产的每个杯子的温度和质量数据。从优化拖拉机跨领域的驾驶方式到协调装配线上的机器人,商业和工业领域有无数种方式可以让更多的数据和更多的连接产生更好的服务和更低的成本。医疗设备和传感器将跟踪和诊断更多疾病。世界拥有的感官信息远远超过我们目前数字化、交流和处理的能力。

The promise of connecting many more devices to cell networks and harvesting data from them may not sound revolutionary. You may not think a 5G network can brew better coffee, but it won’t be long until your coffeemaker is collecting and processing data on the temperature and quality of each cup it produces. There are innumerable ways in business and industry that more data and more connectivity will produce better service and lower cost, from optimizing how tractors drive across fields to coordinating robots on assembly lines. Medical devices and sensors will track and diagnose more conditions. The world has far more sensory information than our current ability to digitize, communicate, and process.

没有比埃隆马斯克的汽车公司特斯拉更好的案例研究表明连接性和计算能力如何将旧产品变成数字化机器。特斯拉的狂热追随和飙升的股价吸引了很多关注,但不太受关注的是,特斯拉也是领先的芯片设计师。该公司聘请了像 Jim Keller 这样的明星半导体设计师来打造一款专门满足其自动驾驶需求的芯片,该芯片采用领先技术制造。早在 2014 年,一些分析师就注意到特斯拉汽车“类似于智能手机。” 该公司经常被比作苹果,后者也设计自己的半导体。与 Apple 的产品一样,Tesla 精心调整的用户体验以及看似轻松地将先进计算集成到 20 世纪产品(汽车)中,都是因为定制设计的芯片才有可能实现。自 1970 年代以来,汽车就采用了简单的芯片。然而,需要专用半导体来管理电源的电动汽车的普及,加上对自动驾驶功能的需求增加,预示着典型汽车中芯片的数量和成本将大幅增加。

There’s no better case study showing how connectivity and computing power will turn old products into digitized machines than Tesla, Elon Musk’s auto company. Tesla’s cult following and soaring stock price have attracted plenty of attention, but what’s less noticed is that Tesla is also a leading chip designer. The company hired star semiconductor designers like Jim Keller to build a chip specialized for its automated driving needs, which is fabricated using leading-edge technology. As early as 2014, some analysts were noting that Tesla cars “resemble a smartphone.” The company has been often compared to Apple, which also designs its own semiconductors. Like Apple’s products, Tesla’s finely tuned user experience and its seemingly effortless integration of advanced computing into a twentieth-century product—a car—are only possible because of custom-designed chips. Cars have incorporated simple chips since the 1970s. However, the spread of electric vehicles, which require specialized semiconductors to manage the power supply, coupled with increased demand for autonomous driving features foretells that the number and cost of chips in a typical car will increase substantially.

汽车只是发送和接收更多数据的能力如何创造对计算能力的更多需求的最突出的例子——在网络“边缘”的设备、蜂窝网络本身以及庞大的数据中心。2017 年前后,随着全球电信公司开始与设备供应商签订建设 5G 网络的合同,中国的华为处于领先地位,提供的设备被业界认为是高品质和有竞争力的价格。华为在 5G 网络建设中的作用似乎比任何其他公司都大,超过了瑞典的爱立信和芬兰的诺基亚,后者是唯一的其他主要手机信号塔设备生产商。

Cars are only the most prominent example of how the ability to send and receive more data will create more demand for computing power—in devices on the “edge” of the network, in the cell network itself, and in vast data centers. Around 2017, as telecom companies around the world began signing contracts with equipment providers to build 5G networks, it emerged that China’s Huawei was in a leading position, offering gear that was perceived by the industry to be high-quality and competitively priced. Huawei looked likely to play a bigger role in the construction of 5G networks than any other company, overtaking Sweden’s Ericcson and Finland’s Nokia, the only other main producers of the equipment on cell towers.

与其竞争对手一样,华为的基站设备内部也有大量的硅。日本报纸《日经亚洲》对华为无线电部门的一项研究发现,该公司严重依赖美国制造的芯片,例如莱迪思半导体(Lattice Semiconductor)的现场可编程门阵列。几年前。德州仪器、Analog Devices、博通和赛普拉斯半导体也设计和制造了华为无线电设备所依赖的芯片。根据这个分析,美国芯片和其他组件占华为每个系统成本的近 30%。但是,主处理器芯片是由华为海思芯片设计部门在国内设计的,并在台积电制造。华为还没有达到技术自给自足的程度。它依靠多家外国芯片公司生产专用半导体,并依靠台积电制造其内部设计的芯片。然而,华为在每个无线电系统中都生产了一些最复杂的电子设备,并了解如何集成所有组件的细节。

Inside Huawei’s equipment on cell towers, like that of its rivals, is a large quantity of silicon. One study of Huawei’s radio units, by the Japanese newspaper Nikkei Asia, found a heavy reliance on U.S.-made chips, like field-programmable gate arrays from Lattice Semiconductor, the Oregon company that Tsinghua Unigroup had bought and then sold a minority stake in several years earlier. Texas Instruments, Analog Devices, Broadcom, and Cypress Semiconductor also designed and built chips that Huawei’s radio gear relied on. According to this analysis, American chips and other components constitute nearly 30 percent of the cost of each Huawei system. However, the main processor chip was designed domestically by Huawei’s HiSilicon chip design arm and fabricated at TSMC. Huawei hadn’t reached technological self-sufficiency. It relied on multiple foreign chip firms to produce specialized semiconductors and on TSMC to fabricate the chips it designed in-house. Yet Huawei produced some of the most complex electronics in each radio system and understood the details of how to integrate all the components.

随着华为的设计部门证明自己是世界级的,不难想象中国芯片设计公司的未来会像硅谷巨头一样成为台积电的重要客户。如果将 2010 年代后期的趋势向前预测,到 2030 年,中国的芯片产业可能会在影响力上与硅谷匹敌。这不会简单地扰乱科技公司和贸易流动。它还将重新设定军事力量的平衡。

With Huawei’s design arm proving itself world-class, it wasn’t hard to imagine a future in which Chinese chip design firms were as important customers of TSMC as Silicon Valley giants. If the trends of the late 2010s were projected forward, by 2030 China’s chip industry might rival Silicon Valley for influence. This wouldn’t simply disrupt tech firms and trade flows. It would also reset the balance of military power.

第 48 章 下一个偏移量

CHAPTER 48 The Next Offset

从成群的自主无人机到网络空间和整个电磁频谱中的隐形战斗,战争的未来将由计算能力决定。美军不再是无可争议的领导者。在精确导弹和全视传感器的保证下,美国拥有无与伦比的进入世界海洋和领空的日子已经一去不复返了。1991 年波斯湾战争后在世界各国国防部间回荡的冲击波——以及对摧毁萨达姆军队的外科手术式打击可能被用来对付世界上任何军队的恐惧——在北京就像一场一种说法是“心理核攻击”。在那次冲突之后的 30 年里,中国将资金投入高科技武器,放弃了毛泽东时代发动低技术人民战争的学说,并接受了未来战斗将依赖于先进传感器、通信和计算。现在,中国正在开发先进战斗部队所需的计算基础设施。

From swarms of autonomous drones to invisible battles in cyberspace and across the electromagnetic spectrum, the future of war will be defined by computing power. The U.S. military is no longer the unchallenged leader. Long gone are the days when the U.S. had unrivaled access to the world’s seas and airspace, guaranteed by precision missiles and all-seeing sensors. The shock waves that reverberated around the world’s defense ministries after the 1991 Persian Gulf War—and the fear that the surgical strikes that had defanged Saddam’s army could be used against any military in the world—was felt in Beijing like a “psychological nuclear attack,” according to one account. In the thirty years since that conflict, China has poured funds into high-tech weaponry, abandoning Mao-era doctrines of waging a low-tech People’s War and embracing the idea that the fights of the future will rely on advanced sensors, communications, and computing. Now China is developing the computing infrastructure an advanced fighting force requires.

北京的目标不仅仅是逐个系统地与美国匹敌,而是发展能够“抵消”美国优势的能力,将五角大楼 1970 年代的概念用于对抗美国。中国部署了一系列有系统地削弱美国优势的武器。中国精准反舰导弹让美国水面舰艇过境极为危险战时的台湾海峡,将美国的海军力量拒之门外。新的防空系统挑战美国在冲突中控制领空的能力。远程陆地攻击导弹威胁着从日本到关岛的美国军事基地网络。中国的反卫星武器威胁要禁用通信和 GPS 网络。中国的网络战能力尚未在战时得到检验,但中国会试图摧毁整个美国的军事系统。同时,在电磁频谱中,中国可能会试图干扰美国的通信和盲目的监视系统,使美军无法看到敌人或与盟友进行通信。

Beijing’s aim isn’t simply to match the U.S. system-by-system, but to develop capabilities that could “offset” American advantages, taking the Pentagon’s concept from the 1970s and turning it against the United States. China has fielded an array of weapons that systematically undermine U.S. advantages. China’s precision anti-ship missiles make it extremely dangerous for U.S. surface ships to transit the Taiwan Strait in a time of war, holding American naval power at bay. New air defense systems contest America’s ability to dominate the airspace in a conflict. Long-range land attack missiles threaten the network of American military bases from Japan to Guam. China’s anti-satellite weapons threaten to disable communications and GPS networks. China’s cyberwar capabilities haven’t been tested in wartime, but the Chinese would try to bring down entire U.S. military systems. Meanwhile, in the electromagnetic spectrum, China might try to jam American communications and blind surveillance systems, leaving the U.S. military unable to see enemies or communicate with allies.

支撑所有这些能力的是中国军界的信念,即战争不仅仅是“信息化”而是“智能化”——这是一种不雅的军事术语,意味着将人工智能应用于武器系统。当然,在过去的半个世纪里,计算能力一直是战争的核心,尽管可以用来支持军事系统的 1 和 0 的数量是几十年前的数百万倍。今天的新鲜事是,美国现在有了一个可信的挑战者。苏联可以逐导弹匹配美国,但不能逐字节匹配。中国认为两者都可以。中国半导体产业的命运不仅仅是商业问题。哪个国家可以生产更多的 1 和 0 也将具有严重的军事优势。

Undergirding all these capabilities is a belief in Chinese military circles that warfare is not simply becoming “informationized” but “intelligentized”—inelegant military jargon that means applying artificial intelligence to weapons systems. Of course, computing power has been central to warfare for the past half century, though the quantity of 1s and 0s that can be harnessed to support military systems is millions of times larger than decades earlier. What’s new today is that America now has a credible challenger. The Soviet Union could match the U.S. missile-for-missile but not byte-for-byte. China thinks it can do both. The fate of China’s semiconductor industry isn’t simply a question of commerce. Whichever country can produce more 1s and 0s will have a serious military advantage, too.

哪些因素将定义这场计算竞赛?2021 年,由谷歌前 CEO 埃里克·施密特 (Eric Sc​​hmidt) 担任主席的一群美国科技和外交政策巨头发布了一份报告,预测“中国可以超越美国成为世界人工智能超级大国。” 中国领导人似乎同意这一点。正如中国军事专家 Elsa Kania 所说,解放军一直在谈论“AI 武器”至少十年,指的是使用“AI 自动追踪、区分和摧毁敌方目标”的系统。习近平本人曾敦促解放军将“加快军事智能化发展”作为国防重点。

What factors will define this computing race? In 2021, a group of American tech and foreign policy grandees chaired by former Google CEO Eric Schmidt released a report predicting that “China could surpass the United States as the world’s AI superpower.” Chinese leaders appear to agree. As China military expert Elsa Kania notes, the PLA has been talking about “AI weapons” for at least a decade, referring to systems that use “AI to pursue, distinguish, and destroy enemy targets automatically.” Xi Jinping himself has urged the PLA to “accelerate the development of military intelligentization” as a defense priority.

军事人工智能的想法让人联想到杀手机器人,但在许多领域应用机器学习可以使军事系统更好。预测性维护——学习何时需要修理机器——已经在帮助飞机保持在空中,船只保持在海上。启用人工智能的潜艇声纳或卫星图像可以更准确地识别威胁。可以更快地设计新的武器系统。炸弹和导弹可以更准确地瞄准,尤其是在移动目标方面。空中、水下和陆地上的自动驾驶汽车已经在学习机动、识别和摧毁对手。并非所有这些都像“人工智能武器”这样的短语所暗示的那样具有革命性。例如,几十年来,我们一直在使用自导式、即发即弃的导弹。但随着武器变得更智能、更自主,它们对计算能力的需求只会增长。

The idea of military AI evokes images of killer robots, but there are many spheres where applying machine learning can make military systems better. Predictive maintenance—learning when machines need to be fixed—is already helping keep planes in the sky and ships at sea. AI-enabled submarine sonars or satellite imagery can identify threats more accurately. New weapons systems can be designed more quickly. Bombs and missiles can be aimed more accurately, especially when it comes to moving targets. Autonomous vehicles in the air, underwater, and on land are already learning to maneuver, identify adversaries, and destroy them. Not all of this is as revolutionary as phrases like “AI weapons” might imply. We’ve had self-guided, fire-and-forget missiles for decades, for example. But as weapons get smarter and more autonomous, their demands for computing power only grow.

不能保证中国一定会赢得开发和部署人工智能系统的竞赛,部分原因是这场“竞赛”不是关于单一技术,而是关于复杂系统。值得记住的是,冷战军备竞赛并不是由第一个向太空发射卫星的国家赢得的。然而,不可否认,中国在人工智能系统方面的能力令人印象深刻。乔治城大学的本·布坎南(Ben Buchanan)指出驾驭人工智能需要数据、算法和计算能力的“三位一体”。除了计算能力,中国的能力可能已经与美国不相上下。

It isn’t guaranteed that China will win the race to develop and deploy systems empowered by artificial intelligence, in part because this “race” isn’t about a single technology but about complex systems. The Cold War arms race, it’s worth remembering, wasn’t won by the first country to shoot a satellite into space. Yet China’s capabilities when it comes to AI systems are undeniably impressive. Georgetown University’s Ben Buchanan has noted that a “triad” of data, algorithms, and computing power are needed to harness AI. With the exception of computing power, China’s capabilities may already equal the United States’.

在访问可以输入人工智能算法的数据类型方面,中国和美国都没有明显的优势。北京的支持者认为,该国的监视状态及其庞大的人口使其能够收集更多数据,尽管收集有关中国民众的数据的能力在军事领域可能没有多大帮助。例如,没有多少关于中国 13 亿公民的网上购物习惯或面部结构的数据可以训练计算机识别潜伏在台湾海峡的潜艇的声音。中国在收集与军事系统相关的数据方面没有任何内在优势。

When it comes to accessing the type of data that can be fed into AI algorithms, neither China nor the U.S. has a clear advantage. Beijing’s boosters argue the country’s surveillance state and its massive population let it collect more data, though the ability to amass data about China’s populace probably doesn’t help much in the military sphere. No amount of data about online shopping habits or the facial structure of all of China’s 1.3 billion citizens will train a computer to recognize the sounds of a submarine lurking in the Taiwan Strait, for example. China doesn’t have any built-in advantages in gathering data relevant to military systems.

在设计巧妙的算法时,很难说一方是否具有优势。以人工智能专家的数量衡量,中国似乎拥有与美国不相上下的能力。专注于中国的智库 MacroPolo 的研究人员发现,全球 29% 的人工智能领先研究人员来自中国,而美国和欧洲分别为 20% 和 18%。然而,这些专家中有很大一部分最终在美国工作,59% 的世界顶级人工智能研究人员。新的签证和旅行限制加上中国在国内留住更多研究人员的努力可能会抵消美国在剥夺地缘政治对手最聪明头脑方面的历史技能。

It’s harder to say whether one side has an advantage when it comes to devising clever algorithms. Measured by the number of AI experts, China appears to have capabilities that are comparable to America’s. Researchers at MacroPolo, a China-focused think tank, found that 29 percent of the world’s leading researchers in artificial intelligence are from China, as opposed to 20 percent from the U.S. and 18 percent from Europe. However, a staggering share of these experts end up working in the U.S., which employs 59 percent of the world’s top AI researchers. The combination of new visa and travel restrictions plus China’s effort to retain more researchers at home may neutralize America’s historical skill at stripping geopolitical rivals of their smartest minds.

在布坎南“三合一”的第三部分,即计算能力中,美国仍然有相当大的领先优势,尽管近年来已经显着削弱。中国仍然严重依赖外国半导体技术——尤其是美国设计、台湾制造的处理器——来进行复杂的计算。依赖外国芯片的不仅仅是中国智能手机和个人电脑。大多数中国数据中心也是如此——这解释了为什么中国如此努力地从 IBM 和 AMD 等公司获取技术。一项中国研究估计,多达例如,运行人工智能工作负载的中国服务器中 95% 的 GPU 是由 Nvidia 设计的。英特尔、赛灵思、AMD 等公司的芯片在中国数据中心中至关重要。即使在最乐观的预测下,中国也需要五年时间才能设计出具有竞争力的芯片及其周围的软件生态系统,而要在国内制造这些芯片的时间要长得多。

In the third part of Buchanan’s “triad,” computing power, the United States still has a substantial lead, though it has eroded significantly in recent years. China is still staggeringly dependent on foreign semiconductor technology—in particular, U.S.-designed, Taiwan-fabricated processors—to undertake complex computation. It isn’t only Chinese smartphones and PCs that rely on foreign chips. So, too, do most Chinese data centers—which explains why the country has tried so hard to acquire technology from companies like IBM and AMD. One Chinese study has estimated that as many as 95 percent of GPUs in Chinese servers running artificial intelligence workloads are designed by Nvidia, for example. Chips from Intel, Xilinx, AMD, and others are crucially important in Chinese data centers. Even under the most optimistic projections, it will be half a decade before China can design competitive chips and the software ecosystem around them, and far longer before it can manufacture these chips domestically.

然而,对于许多中国军事系统而言,获得美国设计、台湾制造的芯片并不难。最近审查了 343 份公开可用的与人工智能相关的中国人民解放军采购合同,由乔治城大学的研究人员发现,只有不到 20% 的合同涉及受美国出口管制的公司。换句话说,中国军方只需购买现成的美国尖端芯片并将其插入军事系统就没有什么困难。乔治城大学的研究人员发现,中国军事供应商甚至在他们的网站上宣传他们使用美国芯片。中国政府备受争议的政策“民用军事融合”是一项将先进民用技术应用于军事系统的努力,看起来它正在发挥作用。如果美国出口限制没有发生重大变化,中国人民解放军只需从硅谷购买即可获得所需的大部分计算能力。

For many Chinese military systems, however, acquiring U.S.-designed, Taiwan-fabricated chips hasn’t been difficult. A recent review of 343 publicly available AI-related People’s Liberation Army procurement contracts, by researchers at Georgetown University, found that less than 20 percent of the contracts involved companies that are subject to U.S. export controls. In other words, the Chinese military has had little difficulty simply buying cutting-edge U.S. chips off-the-shelf and plugging them into military systems. The Georgetown researchers found that Chinese military suppliers even advertise on their websites their use of American chips. The Chinese government’s controversial policy of “Civil Military Fusion,” an effort to apply advanced civilian technology to military systems, looks like it’s working. Absent a major change in U.S. export restrictions, the People’s Liberation Army will acquire much of the computing power it needs by simply buying it from Silicon Valley.

当然,中国人民解放军并不是唯一一个试图将先进计算应用于武器系统的军队。随着中国军队战斗力的增强,五角大楼意识到它需要一个新的战略。2010 年代中期,国防部长查克·哈格尔等官员开始谈到需要新的“抵消”,这引起了比尔·佩里、哈罗德·布朗和安德鲁·马歇尔在 1970 年代克服苏联数量优势的努力。美国今天面临着同样的基本困境:中国可以部署比美国更多的舰船和飞机,尤其是在重要的战区,比如台湾海峡。“我们永远不会试图与我们的对手或我们的竞争对手相提并论,坦克对坦克,飞机对飞机,人对人,”前国防部副部长鲍勃·沃克(Bob Work)宣称,他是这个新抵消的知识分子教父,清晰地呼应了 1970 年代后期的逻辑。换句话说,美军只有成功决定性的技术优势。

Of course, the People’s Liberation Army isn’t the only military trying to apply advanced computing to weapons systems. As the fighting power of China’s military has grown, the Pentagon has realized it needs a new strategy. In the mid-2010s, officials like Secretary of Defense Chuck Hagel began speaking about a need for a new “offset,” evoking the effort of Bill Perry, Harold Brown, and Andrew Marshall during the 1970s to overcome the USSR’s quantitative advantage. The U.S. faces the same basic dilemma today: China can deploy more ships and planes than the U.S., especially in theaters that matter, like the Taiwan Strait. “We will never try to match our opponents or our competitors tank for tank, plane for plane, person for person,” declared Bob Work, the former deputy defense secretary who is the intellectual godfather of this new offset, in a clear echo of the logic of the late 1970s. The U.S. military will only succeed, in other words, if it has a decisive technological advantage.

这种技术优势会是什么样子?沃克认为,1970 年代的偏移是由“数字微处理器、信息技术、新传感器、隐形”驱动的。这一次,它将是“人工智能(AI)和自治的进步”。美国军方已经在部署第一代新型自动驾驶汽车,例如 Saildrone,这是一种无人驾驶风帆冲浪者,可以在跟踪潜艇或拦截对手通信的同时花费数月时间在海洋中漫游。这些设备的成本只是典型海军舰艇的一小部分,让军事领域中的许多设备得以使用,并为跨世界海洋的传感器和通信提供平台。自主水面舰艇、飞机和潜艇也在开发和部署中。这些自主平台将需要人工智能来指导它们并做出决策。

What will this technological advantage look like? The 1970s offset was driven by “digital microprocessors, information technologies, new sensors, stealth,” Work has argued. This time, it will be “advances in Artificial Intelligence (AI) and autonomy.” The U.S. military is already fielding the first generation of new autonomous vehicles, like Saildrone, an unmanned windsurfer that can spend months roving the oceans while tracking submarines or intercepting adversaries’ communications. These devices cost a tiny fraction of a typical Navy ship, letting the military field many of them and providing platforms for sensors and communications across the world’s oceans. Autonomous surface ships, planes, and submarines are also being developed and deployed. These autonomous platforms will require artificial intelligence to guide them and make decisions. The more computing power that can be put on board, the smarter decisions they’ll make.

DARPA 开发了使 1970 年代偏移成为可能的技术;现在它正在设计能够在战争中实现新的计算转换的系统。DARPA 领导人设想“分布在战场上的计算机可以相互通信和协调”,从最大的海军舰艇到最小的无人机。挑战不仅仅是将计算能力嵌入到单个设备中,例如导弹,而是将战场上的数千台设备联网,让它们共享数据并使机器能够做出更多决策。DARPA 资助了“人机协作”,例如,设想一架有人驾驶的战斗机与几架自主无人机一起飞行,这些无人机是人类飞行员的额外眼睛和耳朵。

DARPA developed the technology that made the 1970s offset possible; now it’s devising systems that promise new computing-enabled transformations in warfare. DARPA leaders envision “computers distributed across the battlespace that can all communicate and coordinate with one another,” from the largest naval ship to the tiniest drone. The challenge isn’t simply to embed computing power in a single device, like a guided missile, but to network thousands of devices across a battlefield, letting them share data and putting machines in a position to make more decisions. DARPA has funded research programs on “human-machine teaming,” envisioning, for example, a piloted fighter jet flying alongside several autonomous drones that are an additional set of eyes and ears for the human pilot.

正如冷战是由环绕美国导弹制导计算机的电子决定的一样,未来的战斗可能会在电磁频谱中决定。世界军队对电子传感器和通信的依赖程度越高,他们就越需要为获取发送信息或检测和跟踪对手所需的频谱空间而战。我们只是瞥见了战时电磁频谱操作的样子。例如,俄罗斯在对乌克兰的战争中使用了各种雷达和信号干扰器。据报道,俄罗斯政府还阻碍了总统弗拉基米尔·普京公务旅行周围的 GPS 信号,这可能是一种安全措施。并非巧合的是,DARPA 正在研究不依赖 GPS 信号或卫星的替代导航系统,即使 GPS 系统出现故障,美国导弹也能击中目标。

Just as the Cold War was decided by electrons zipping around the guidance computers of American missiles, the fights of the future may be decided in the electromagnetic spectrum. The more the world’s militaries rely on electronic sensors and communication, the more they’ll have to battle for access to the spectrum space needed to send messages or to detect and track adversaries. We’ve only had a glimpse of what wartime electromagnetic spectrum operations will look like. For example, Russia has used a variety of radar and signals jammers in its war against Ukraine. The Russian government also reportedly obstructs GPS signals around President Vladimir Putin’s official travel, perhaps as a security measure. Not coincidentally, DARPA is researching alternative navigation systems that aren’t reliant on GPS signals or satellites, to enable American missiles to hit their targets even if GPS systems are down.

电磁频谱之战将是一场由半导体进行的无形之争。雷达、干扰和通信都由复杂的射频芯片和数模转换器管理,它们调制信号以利用开放频谱空间,向特定方向发送信号,并试图迷惑对手的传感器。同时,强大的数字芯片将在雷达或干扰器中运行复杂的算法,评估接收到的 289 个信号,并在几毫秒内决定发送哪些信号。危在旦夕军队的观察和交流能力。如果设备无法确定它们的位置或前进方向,那么自主无人机将不值钱。

The battle for the electromagnetic spectrum will be an invisible struggle conducted by semiconductors. Radar, jamming, and communications are all managed by complex radio frequency chips and digital-analog converters, which modulate signals to take advantage of open spectrum space, send signals in a specific direction, and try to confuse adversaries’ sensors. Simultaneously, powerful digital chips will run complex algorithms inside a radar or jammer that assess the 289signals received and decide what signals to send out in a matter of milliseconds. At stake is a military’s ability to see and to communicate. Autonomous drones won’t be worth much if the devices can’t determine where they are or where they’re heading.

未来的战争将比以往任何时候都更加依赖芯片——运行人工智能算法的强大处理器、处理数据的大内存芯片、感知和产生无线电波的完美调谐模拟芯片。2017 年,DARPA 启动了一个名为电子复兴计划,以帮助建立下一波军事相关芯片技术。在某些方面,DARPA 对芯片的重新兴趣自然源于其历史。它资助了加州理工学院的 Carver Mead 等先驱学者,并促进了对芯片设计软件、新光刻技术和晶体管结构。

The warfare of the future will be more reliant than ever on chips—powerful processors to run AI algorithms, big memory chips to crunch data, perfectly tuned analog chips to sense and produce radio waves. In 2017, DARPA launched a new project called the Electronics Resurgence Initiative to help build the next wave of militarily relevant chip technology. In some ways, DARPA’s renewed interest in chips stems naturally from its history. It funded pioneering scholars like Caltech’s Carver Mead and catalyzed research into chip design software, new lithography techniques, and transistor structures.

然而,DARPA 和美国政府发现塑造芯片行业的未来比以往任何时候都更加困难。DARPA 的预算是每年几十亿美元,低于大多数业内最大公司的研发预算。当然,DARPA 在遥远的研究想法上花费更多,而英特尔和高通等公司将大部分资金花在距离成果只有几年的项目上。然而,总体而言,美国政府购买的全球芯片份额比以往任何时候都小。美国政府购买了飞兆半导体和德州仪器公司在 1960 年代初期生产的几乎所有早期集成电路。到 1970 年代,这个数字已经下降到10-15%。现在它约占美国芯片市场的 2%。作为芯片的买家,苹果 CEO 蒂姆库克对这个行业的影响力超过了当今五角大楼的任何一位官员。

Yet DARPA and the U.S. government have found it harder than ever to shape the future of the chip industry. DARPA’s budget is a couple billion dollars per year, less than the R&D budgets of most of the industry’s biggest firms. Of course, DARPA spends a lot more on far-out research ideas, whereas companies like Intel and Qualcomm spend most of their money on projects that are only a couple years from fruition. However, the U.S. government in general buys a smaller share of the world’s chips than ever before. The U.S. government bought almost all the early integrated circuits that Fairchild and Texas Instruments produced in the early 1960s. By the 1970s, that number had fallen to 10−15 percent. Now it’s around 2 percent of the U.S. chip market. As a buyer of chips, Apple CEO Tim Cook has more influence on the industry than any Pentagon official today.

制造半导体的成本如此之高,以至于五角大楼都无法在内部进行。国家安全局曾经在其位于马里兰州米德堡的总部设有芯片工厂。然而,在 2000 年代,政府认为按照摩尔定律规定的节奏进行升级太昂贵了。今天甚至设计了一个尖端的芯片——这可能会花费很多几亿美元——对于除了最重要的项目之外的所有项目来说都太昂贵了。

Making semiconductors is so expensive that even the Pentagon can’t afford to do it in-house. The National Security Agency used to have a chip fab at its headquarters in Maryland’s Fort Meade. In the 2000s, however, the government decided it was too expensive to keep upgrading per the cadence dictated by Moore’s Law. Today even designing a leading-edge chip—which can cost several hundred million dollars—is too expensive for all but the most important projects.

美国军方和政府的间谍机构都将其芯片的生产外包给“值得信赖的代工厂”。对于美国拥有世界级能力的许多类型的模拟或射频芯片来说,这相对简单。然而,当涉及到逻辑芯片时,这就造成了两难境地。英特尔的生产能力只是落后于领先优势,尽管该公司主要为自己的个人电脑和服务器业务生产芯片。与此同时,台积电和三星在台湾和韩国保留了最先进的制造能力。很大一部分芯片组装和封装也发生在亚洲。随着国防部试图使用更多现成的组件来降低成本,它将从国外购买更多的设备。

Both the U.S. military and the government’s spy agencies outsource the production of their chips to “trusted foundries.” This is relatively straightforward for many types of analog or radio frequency chips, where the U.S. has world-class capabilities. When it comes to logic chips, though, this poses a dilemma. Intel’s production capabilities are just behind the leading edge, though the company mostly produces chips for its own PC and server businesses. TSMC and Samsung, meanwhile, keep their most cutting-edge fabrication capabilities in Taiwan and South Korea. And a large share of chip assembly and packaging also takes place in Asia. As the Defense Department tries to use more off-the-shelf components to reduce cost, it will buy even more devices from abroad.

军方担心在国外制造或组装的芯片更容易被篡改,添加后门或写入错误。然而,即使是在国内设计和生产的芯片也可能存在意想不到的漏洞。2018 年,研究人员在英特尔广泛使用的微处理器架构中发现了两个基本错误,称为Spectre 和 Meltdown 可以复制密码等数据,这是一个巨大的安全漏洞。据《华尔街日报》报道,英特尔首先向包括中国科技公司在内的客户披露了该漏洞,在通知美国政府之前,这一事实只会加剧五角大楼官员对其对芯片行业影响力下降的担忧。

The military worries that chips fabricated or assembled abroad are more susceptible to tampering, with back doors added or errors written in. However, even chips designed and produced domestically can have unintended vulnerabilities. In 2018, researchers discovered two fundamental errors in Intel’s widely used microprocessor architecture called Spectre and Meltdown, which enabled the copying of data such as passwords—a huge security flaw. According to the Wall Street Journal, Intel first disclosed the flaw to customers, including Chinese tech companies, before notifying the U.S. government, a fact that only intensified Pentagon officials’ concern about their declining influence over the chip industry.

DARPA 正在投资于可以保证芯片不可篡改或验证它们是否完全按照预期制造的技术。军方可以指望 TI 等公司在陆上设计、制造和组装尖端模拟和数字电子产品的日子已经一去不复返了。今天根本没有办法避免从国外购买一些东西——并且从台湾购买许多东西。因此,DARPA 押注于技术以实现微电子学的“零信任”方法:不信任任何东西并验证一切,通过植入芯片上的微型传感器等技术,可以检测到修改它的努力。

DARPA is investing in technology that can guarantee chips are tamper-free or to verify they’re manufactured exactly as intended. Long gone are the days when the military could count on firms like TI to design, manufacture, and assemble cutting-edge analog and digital electronics all onshore. Today there’s simply no way to avoid buying some things from abroad—and buying many from Taiwan. So DARPA’s betting on technology to enable a “zero trust” approach to microelectronics: trust nothing and verify everything, via technologies like tiny sensors implanted on a chip that can detect efforts to modify it.

所有这些努力利用微电子来刺激新的“抵消”并重新建立对中国和俄罗斯的决定性军事优势,但是,假设美国将保持其在芯片方面的领先地位。现在看来,这是一个冒险的赌注。在“跑得更快”战略的时代,美国在芯片制造过程的某些环节落后,尤其是在制造先进逻辑芯片方面对台湾的依赖日益增加。三年来一直是美国芯片冠军的英特尔,现在显然已经步履蹒跚。许多业内人士认为,它已经决定性地落后了。与此同时,中国正在向其芯片产业投入数十亿美元,同时迫使外国公司交出敏感技术。对于每一家大型芯片公司来说,中国消费市场是比美国政府重要得多的客户。

All these efforts to use microelectronics to spur a new “offset” and reestablish a decisive military advantage over China and Russia, however, assume the U.S. will keep its lead in chips. That’s now looking like a risky bet. The era of the “run faster” strategy saw the U.S. fall behind in certain segments of the chipmaking process, most notably in the growing dependence on Taiwan for building advanced logic chips. Intel, which for three decades had been America’s chip champion, has now very clearly stumbled. Many people in the industry think it has fallen decisively behind. Meanwhile, China is pouring billions of dollars into its chip industry while pressuring foreign companies to turn over sensitive technology. For every major chip firm, the Chinese consumer market is far more important a customer than the U.S. government.

北京努力获取先进技术、美国和中国电子工业之间的深度互联以及两国相互依赖台湾制造都引发了质疑。美国已经跑得慢了。它是现在将其军队的未来押注在其主导地位正在下滑的技术上。五角大楼负责该问题的官员马特·特平认为:“这种以抵消方式推进的想法,如果中国人和我们一起在车里。”

Beijing’s efforts to acquire advanced technology, the deep interconnections between the U.S. and Chinese electronics industries, and the two countries’ mutual reliance on fabrication in Taiwan all raise questions. America was already running slower. It’s now betting the future of its military on a technology over which its dominance is slipping. “This idea of pulling ahead with an offset,” argues Matt Turpin, an official who worked on the issue at the Pentagon, “is nearly impossible if the Chinese are in the car with us.”

习近平宣布,“发动突击”。中国领导人已将其对外国芯片制造商的依赖视为一个关键漏洞。他们制定了一项计划,通过收购外国芯片制造商、窃取他们的技术以及向中国芯片公司提供数十亿美元的补贴来改造世界芯片产业。中国人民解放军现在指望这些努力来帮助它规避美国的限制,尽管它仍然可以合法地购买许多美国芯片来追求“军事智能化”。就五角大楼而言,在承认中国的军事现代化已经缩小了两个超级大国军队之间的差距,特别是在中国沿海有争议的海域之后,它已经开始了自己的抵消。台湾不仅仅是两国军方押宝的先进芯片的来源地。

“Call forth the assault,” Xi Jinping declared. China’s leaders have identified their reliance on foreign chipmakers as a critical vulnerability. They’ve set out a plan to rework the world’s chip industry by buying foreign chipmakers, stealing their technology, and providing billions of dollars of subsidies to Chinese chip firms. The People’s Liberation Army is now counting on these efforts to help it evade U.S. restrictions, though it can still buy legally many U.S. chips in its pursuit of “military intelligentization.” For its part, the Pentagon has launched its own offset, after admitting that China’s military modernization has closed the gap between the two superpowers’ militaries, especially in the contested waters off China’s coast. Taiwan isn’t simply the source of the advanced chips that both countries’ militaries are betting on. It’s also the most likely future battleground.

第八部分 芯片扼流圈

PART VIII THE CHIP CHOKE

第 49 章 “我们竞争的一切”

CHAPTER 49 “Everything We’re Competing On”

ntel 的 CEO Brian Krzanich 无法掩饰他对中国努力抢占世界芯片行业更大份额的焦虑。作为美国芯片行业贸易组织半导体行业协会 2015 年的主席,科再奇的任务是与美国政府官员打交道。通常这意味着要求减税或减少监管。这一次,话题不同了:说服美国政府对中国的巨额半导体补贴做点什么。美国的芯片公司都陷入了同样的困境。对于几乎所有美国半导体公司来说,中国都是一个至关重要的市场,要么是因为这些公司直接向中国客户销售产品,要么是因为它们的芯片在中国组装成智能手机或电脑。北京的强硬手段迫使美国芯片公司对中国的补贴保持沉默,

Intel’s CEO Brian Krzanich couldn’t hide his anxiety about China’s push to seize a bigger share of the world’s chip industry. As chairman in 2015 of the Semiconductor Industry Association, the U.S. chip industry’s trade group, Krzanich was tasked with hobnobbing with U.S. government officials. Usually this meant asking for tax cuts or reduced regulation. This time, the topic was different: convincing the U.S. government to do something about China’s massive semiconductor subsidies. America’s chip firms were all caught in the same bind. China was a crucial market for almost every U.S. semiconductor firm, either because these firms sold directly to Chinese customers or because their chips were assembled into smartphones or computers in China. Beijing’s strong-arm methods forced U.S. chip firms to stay silent about China’s subsidies, even though the Chinese government had adopted a formal policy of trying to cut them out of China’s supply chain.

奥巴马政府官员习惯于钢铁或太阳能电池板等行业对中国的抱怨。高科技本应是美国的专长,是它具有竞争优势的领域。因此,当高级政府官员察觉到与科再奇会面时,“他眼中明显的恐惧感”,他们很担心。英特尔的当然,CEO 有很长的偏执历史。但现在公司和整个美国芯片行业比以往任何时候都更有理由担心。中国已将美国太阳能电池板制造业赶出市场。它不能在半导体中做同样的事情吗?“这个 2500 亿美元的巨额基金是会埋葬我们,”一位奥巴马官员担心,提到中国中央和地方政府承诺支持本土芯片制造商的补贴。

Obama administration officials were used to complaints about China from industries like steel or solar panels. High tech was supposed to be America’s specialty, a sphere where it had a competitive advantage. So when senior administration officials perceived a “palpable sense of fear in his eyes” when meeting with Krzanich, they were worried. Intel’s CEOs had a long history of paranoia, of course. But now there was more reason than ever for the company, and the entire U.S. chip industry, to be worried. China had driven U.S. solar panel manufacturing out of business. Couldn’t it do the same in semiconductors? “This massive $250 billion fund is going to bury us,” one Obama official worried, referencing the subsidies China’s central and local governments have promised to support homegrown chipmakers.

到 2015 年左右,从美国政府的深处,齿轮开始慢慢转变。政府的贸易谈判代表认为中国的芯片补贴公然违反了国际协议。五角大楼紧张地注视着中国将计算能力应用于新武器系统的努力。情报机构和司法部发现了更多中国政府与其产业勾结排挤美国芯片公司的证据。然而,美国科技政策的两大支柱——拥抱全球化和“跑得更快”——已经根深蒂固,不仅来自行业的游说,还来自华盛顿的知识分子共识。此外,华盛顿的大多数人几乎不知道半导体是什么。一位参与这项工作的人士回忆说,奥巴马政府在半导体方面进展缓慢,不认为芯片是一个重要问题。

By around 2015, from deep in the U.S. government, gears slowly began to shift. The government’s trade negotiators saw China’s chip subsidies as a flagrant violation of international agreements. The Pentagon nervously watched China’s efforts to apply computing power to new weapons systems. The intelligence agencies and Justice Department unearthed more evidence of collusion between China’s government and its industries to push out American chip firms. Yet the twin pillars of American tech policy—embracing globalization and “running faster”—were deeply ingrained, not only by the industry’s lobbying, but also by Washington’s intellectual consensus. Moreover, most people in Washington barely knew what a semiconductor was. The Obama administration moved slowly on semiconductors, one person involved in the effort recalled, because many senior officials simply didn’t see chips as an important issue.

因此,直到奥巴马政府的最后几天,政府才开始采取行动。2016 年底,即当年总统大选前六天,商务部长彭尼普利兹克在华盛顿发表了关于半导体的高调演讲,宣称“半导体技术仍然是美国独创性的核心特征和我们经济增长的驱动力,这一点势在必行。我们不能放弃我们的领导权。” 她将中国视为核心挑战,谴责“不公平的贸易行为和大规模、非市场化的国家干预”,并引用“中国基于政府利益而非商业目标收购公司和技术的新尝试”清华紫光集团的收购狂潮。

It wasn’t until the final days of the Obama administration, therefore, that the government began to act. In late 2016, six days before that year’s presidential election, Commerce Secretary Penny Pritzker gave a high-profile address in Washington on semiconductors, declaring it “imperative that semiconductor technology remains a central feature of American ingenuity and a driver of our economic growth. We cannot afford to cede our leadership.” She identified China as the central challenge, condemning “unfair trade practices and massive, non-market-based state intervention” and cited “new attempts by China to acquire companies and technology based on their government’s interest—not commercial objectives,” an accusation driven by Tsinghua Unigroup’s acquisition spree.

然而,由于奥巴马政府所剩时间不多,普利兹克无能为力。相反,政府的温和目标是开始一场讨论——它希望——即将上任的希拉里·克林顿政府能够继续推进。普利兹克还下令商务部对半导体供应链进行研究,并承诺“利用一切机会向中国领导人明确表示,我们不会接受旨在侵占该行业的 1500 亿美元产业政策。” 但谴责中国的补贴很容易。让他们停下来要困难得多。

With little time left in the Obama administration, however, there wasn’t much Pritzker could do. Rather, the administration’s modest goal was to start a discussion that—it hoped—the incoming Hillary Clinton administration would carry forward. Pritzker also ordered the Commerce Department to conduct a study of the semiconductor supply chain and promised to “make clear to China’s leaders at every opportunity that we will not accept a $150 billion industrial policy designed to appropriate this industry.” But it was easy to condemn China’s subsidies. It was far harder to make them stop.

大约在同一时间,白宫委托一群半导体高管和学者研究该行业的未来。他们在奥巴马卸任前几天发布了一份报告,敦促美国在现有战略上加倍努力。它的主要建议是:“通过跑得更快来赢得比赛”——这些建议可以从 1990 年代复制和粘贴。不断创新的需要显然很重要。摩尔定律的延续是竞争的必然。但在华盛顿认为它“跑得更快”的几十年里,它的对手已经扩大了他们的市场份额,而整个世界已经变得可怕地依赖于少数易受攻击的瓶颈,尤其是台湾。

Around the same time, the White House commissioned a group of semiconductor executives and academics to study the future of the industry. They issued a report days before Obama left office, which urged the U.S. to double down on its existing strategy. Its primary recommendation was: “win the race by running faster”—advice that could have been copied and pasted from the 1990s. The need to keep innovating was obviously important. The continuation of Moore’s Law was a competitive necessity. But during the decades Washington thought it was “running faster,” its adversaries had grown their market share while the entire world had become frighteningly reliant on a handful of vulnerable choke points, in particular Taiwan.

在华盛顿和芯片行业,几乎每个人都喝过自己关于全球化的“酷爱”。报纸和学者都报道说,全球化实际上是“全球性的”,技术传播是不可阻挡的,其他国家的先进技术能力符合美国的利益,即使它们不是,也没有什么可以阻止技术进步。“在半导体行业全球化的世界里,单边行动越来越无效,”奥巴马政府的半导体报告宣称。“政策原则上可以减缓技术的传播,但不能阻止传播。” 这些说法都没有证据支持;他们只是被假定为真实的。然而,芯片制造的“全球化”并没有发生;“台湾化”了。技术尚未扩散。美国科技政策受制于容易被视为错误的全球化陈词滥调。

In Washington and in the chip industry, almost everyone had drunk their own Kool-Aid about globalization. Newspapers and academics alike reported that globalization was in fact “global,” that technological diffusion was unstoppable, that other countries’ advancing technological capabilities were in the U.S. interest, and that even if they weren’t, nothing could halt technological progress. “Unilateral action is increasingly ineffective in a world where the semiconductor industry is globalized,” the Obama administration’s semiconductor report declared. “Policy can, in principle, slow the diffusion of technology, but it cannot stop the spread.” Neither of these claims was backed by evidence; they were simply assumed to be true. However, “globalization” of chip fabrication hadn’t occurred; “Taiwanization” had. Technology hadn’t diffused. It was monopolized by a handful of irreplaceable companies. American tech policy was held hostage to banalities about globalization that were easily seen to be false.

美国在制造、光刻和其他领域的技术领先地位已经消失,因为华盛顿确信公司应该竞争,而政府应该只是提供一个公平的竞争环境。如果每个国家都同意,自由放任制度就会奏效。许多政府,尤其是亚洲的政府,都深度参与支持其芯片产业。然而,美国官员发现更容易忽视其他国家抢占宝贵芯片产业份额的努力,而是选择鹦鹉学舌地谈论自由贸易和公开竞争。与此同时,美国的地位正在削弱。

America’s technological lead in fabrication, lithography, and other fields had dissipated because Washington convinced itself that companies should compete but that governments should simply provide a level playing field. A laissez-faire system works if every country agrees to it. Many governments, especially in Asia, were deeply involved in supporting their chip industries. However, U.S. officials found it easier to ignore other countries’ efforts to grab valuable chunks of the chip industry, instead choosing to parrot platitudes about free trade and open competition. Meanwhile, America’s position was eroding.

在华盛顿和硅谷的礼貌公司里,简单地重复多边主义、全球化和创新之类的词更容易,这些概念太空洞而不会冒犯任何当权者。芯片行业本身——非常害怕激怒中国或台积电——将其大量游说资源用于重复有关该行业已变得如何“全球化”的虚假陈词滥调。这些概念与在美国单极时期指导两党官员的自由主义国际主义精神自然契合。当每个人都假装合作是双赢的时候,与外国公司和政府的会面更加愉快。所以华盛顿一直告诉自己,美国跑得更快,盲目地忽视美国地位的恶化,中国能力的上升,

In polite company in Washington and Silicon Valley, it was easier simply to repeat words like multilateralism, globalization, and innovation, concepts that were too vacuous to offend anyone in a position of power. The chip industry itself—deeply fearful of angering China or TSMC—put its considerable lobbying resources behind repeating false platitudes about how “global” the industry had become. These concepts fit naturally with the liberal internationalist ethos that guided officials of both political parties amid America’s unipolar moment. Meetings with foreign companies and governments were more pleasant when everyone pretended that cooperation was win-win. So Washington kept telling itself that the U.S. was running faster, blindly ignoring the deterioration in the U.S. position, the rise in China’s capabilities, and the staggering reliance on Taiwan and South Korea, which grew more conspicuous every year.

然而,在美国政府的深处,国家安全官僚机构开始采取不同的观点。这部分政府是为了偏执而付出的,因此安全官员对中国的科技行业持怀疑态度,对政府持更愤世嫉俗的态度也就不足为奇了。许多官员担心中国对世界关键技术系统的影响力正在增长。他们还假设中国将利用其作为世界主要电子产品制造商的地位来插入后门并更有效地进行间谍活动,就像美国所做的那样几十年。设计未来武器的五角大楼官员开始意识到他们对半导体的依赖程度。与此同时,官员们专注于电信基础设施,他们担心美国盟友从欧洲和美国购买的电信设备减少,而从中兴通讯和华为等中国公司购买的电信设备更多。

Deep in the U.S. government, however, the national security bureaucracy was coming to adopt a different view. This part of the government is paid to be paranoid, so it’s no surprise security officials viewed China’s tech industry more skeptically and its government more cynically. Many officials worried that China’s leverage over the world’s critical technology systems was growing. They also presumed China would use its position as the world’s key manufacturer of electronics to insert back doors and to spy more effectively, just as the U.S. had done for decades. Pentagon officials devising weapons of the future began to realize how reliant they’d be on semiconductors. Officials focused on telecom infrastructure, meanwhile, worried that U.S. allies were buying less telecom equipment from Europe and the U.S. and more from Chinese firms like ZTE and Huawei.

多年来,美国情报部门一直对华为涉嫌与中国政府有联系表示担忧,尽管直到 2010 年代中期,该公司及其规模较小的同行中兴通讯才开始引起公众关注。两家公司都出售相互竞争的电信设备;中兴通讯是国有的,而华为是私有的,但被美国官员指控与政府关系密切。两家公司都有花了数十年时间与有关他们贿赂多个国家的官员以赢得合同的指控作斗争。而在 2016 年,也就是奥巴马政府执政的最后一年,两人都是被指控向伊朗和朝鲜供应商品违反了美国的制裁。

U.S. intelligence had voiced concerns about Huawei’s alleged links to the Chinese government for many years, though it was only in the mid-2010s that the company and its smaller peer, ZTE, started attracting public attention. Both companies sold competing telecom equipment; ZTE was state-owned, while Huawei was private but was alleged by U.S. officials to have close ties with the government. Both companies had spent decades fighting allegations that they’d bribed officials in multiple countries to win contracts. And in 2016, during the final year of the Obama administration, both were accused of violating U.S. sanctions by supplying goods to Iran and North Korea.

奥巴马政府考虑对中兴通讯实施金融制裁,这将切断该公司进入国际银行系统的渠道,但取而代之的是选择在 2016 年通过限制美国公司向其出售产品来惩罚该公司。例如,此类出口管制以前主要用于针对军事目标,以阻止向为伊朗导弹计划提供零部件的公司转让技术。但商务部也拥有广泛的权力来禁止民用技术的出口。中兴在其系统中高度依赖美国组件——尤其是美国芯片。然而,2017年3月,在威胁限制实施之前,该公司与美国政府签署认罪协议并支付了罚款,因此取消了出口限制在他们采取武力之前。几乎没有人知道禁止一家大型中国科技公司购买美国芯片的举动会有多么激烈。

The Obama administration considered imposing financial sanctions on ZTE, which would have severed the company’s access to the international banking system, but instead opted to punish the company in 2016 by restricting U.S. firms from selling to it. Export controls like this had previously been used mostly against military targets, to stop the transfer of technology to companies supplying components to Iran’s missile program, for example. But the Commerce Department had broad authority to prohibit the export of civilian technologies, too. ZTE was highly reliant on American components in its systems—above all, American chips. However, in March 2017, before the threatened restrictions were implemented, the company signed a plea deal with the U.S. government and paid a fine, so the export restrictions were removed before they’d taken force. Hardly anyone understood just how drastic a move it would have been to ban a major Chinese tech company from buying U.S. chips.

中兴的认罪协议是在特朗普政府上任之际签署的。特朗普一再抨击中国“敲诈我们”,但他对政策细节不感兴趣,对技术也不感兴趣。他的重点是在贸易和关税方面,彼得·纳瓦罗和罗伯特·莱特希泽等他的官员试图减少双边贸易逆差和减缓离岸外包,但大多以失败告终。然而,在国家安全委员会上,由前记者和海军陆战队员马特·波廷格(Matt Pottinger)领导的少数谨慎官员远离政治风头,最终升任特朗普的副国家安全顾问,他们正在改变美国的对华政策,摆脱几十年的技术政策在此过程中。NSC 中的对华鹰派关注的是北京的地缘政治议程及其技术基础,而不是关税。他们认为美国的地位已经危险地削弱了,应该归咎于华盛顿的不作为。“这真的很重要,你无能为力。”

ZTE’s plea deal was signed just as the Trump administration took office. Trump repeatedly attacked China for “ripping us off,” but he had little interest in policy details and none in technology. His focus was on trade and tariffs, where his officials like Peter Navarro and Robert Lighthizer tried and mostly failed to reduce the bilateral trade deficit and slow offshoring. Far from the political limelight, however, on the National Security Council, a handful of discreet officials led by Matt Pottinger, a former journalist and Marine, who eventually rose to become Trump’s deputy national security advisor, were transforming America’s policy toward China, casting off several decades of technology policy in the process. Rather than tariffs, the China hawks on the NSC were fixated on Beijing’s geopolitical agenda and its technological foundation. They thought America’s position had weakened dangerously and Washington’s inaction was to blame. “This is really important,” one Trump appointee reported an Obama official telling him during the presidential transition, regarding China’s technological advances, “but there’s nothing you can do.”

新政府的中国团队不同意。正如一位高级官员所说,他们得出的结论是,“我们在 21 世纪竞争的一切……都建立在掌握半导体的基石之上。” 他们认为,无所作为不是一个可行的选择。“跑得更快”也不是——他们认为这是无所作为的代码。一位 NSC 官员表示:“如果我们跑得更快,那将是一件好事,”但由于中国“在迫使技术更替方面的巨大影响力”,该战略没有奏效。新的国家安全委员会对技术政策采取了更加好斗的零和方法。从财政部投资审查部门的官员到管理五角大楼军事系统供应链的官员,作为与中国打交道的战略的一部分,政府开始将重点放在半导体上。

The new administration’s China team didn’t agree. They concluded, as one senior official put it, “that everything we’re competing on in the twenty-first century… all of it rests on the cornerstone of semiconductor mastery.” Inaction wasn’t a viable option, they believed. Nor was “running faster”—which they saw as code for inaction. “It would be great for us to run faster,” one NSC official put it, but the strategy didn’t work because of China’s “enormous leverage in forcing the turnover of technology.” The new NSC adopted a much more combative, zero-sum approach to technology policy. From the officials in the Treasury Department’s investment screening unit to those managing the Pentagon’s supply chains for military systems, key elements of the government began focusing on semiconductors as part of their strategy for dealing with China.

这让半导体行业的领导者深感不安。他们希望得到政府的帮助,但又害怕中国的报复。芯片行业会欣然接受降低税收或减少监管,这两者都会使在美国开展业务更具吸引力,但它不想不得不改变其跨国业务模型。硅谷的许多人厌恶特朗普并没有帮助。英特尔首席执行官布赖恩Krzanich 在特朗普作为候选人时同意为特朗普举行筹款活动后遭到强烈反对。然后,在加入了由白宫召集的咨询委员会后,科再奇后来辞去了该委员会的职务。即使行业高管忽视了特朗普的国内政策,他的反复无常也让他成为一个有问题的盟友。通过推文宣布关税从来都不是让 CEO 印象深刻的策略。

This made the semiconductor industry’s leaders deeply uncomfortable. They wanted the government’s help but feared Chinese retaliation. The chip industry would happily accept lower taxes or reduced regulation, both of which would make doing business in the U.S. more attractive, but it didn’t want to have to change its multinational business model. It didn’t help matters that many in Silicon Valley detested Trump. Intel’s CEO Brian Krzanich faced a backlash after agreeing to hold a fundraiser for Trump when he was a candidate. Then, after joining an advisory council convened by the White House, Krzanich later resigned from it. Even when industry executives overlooked Trump’s domestic policies, his volatility made him a problematic ally. Announcing tariffs via tweet was never a tactic that would impress CEOs.

然而,来自芯片行业的信息并不比特朗普白宫自相矛盾的泄密更加连贯。在公开场合,半导体首席执行官及其游说者敦促新政府与中国合作,并鼓励其遵守贸易协定。私下里,他们承认这种策略是没有希望的,并担心国家支持的中国竞争对手会以牺牲他们为代价抢占市场份额。整个芯片行业都依赖于对中国的销售——无论是英特尔这样的芯片制造商、高通这样的无晶圆厂设计师,还是应用材料公司这样的设备制造商。一位美国半导体高管向一位白宫官员讽刺地总结道:“我们的根本问题是,我们的第一大客户是我们的第一大竞争对手。”

However, the messages coming from the chip industry weren’t any more coherent than the contradictory leaks from the Trump White House. Publicly, semiconductor CEOs and their lobbyists urged the new administration to work with China and encourage it to comply with trade agreements. Privately, they admitted this strategy was hopeless and feared that state-supported Chinese competitors would grab market share at their expense. The entire chip industry depended on sales to China—be it chipmakers like Intel, fabless designers like Qualcomm, or equipment manufacturers like Applied Materials. One U.S. semiconductor executive wryly summed things up to a White House official: “Our fundamental problem is that our number one customer is our number one competitor.”

国家安全委员会中的对华鹰派认为,美国的半导体产业需要自救。任凭股东和市场力量一时兴起,芯片公司将慢慢将员工、技术和知识产权转移到中国,直到硅谷被掏空。对华鹰派认为,美国需要更强有力的出口管制制度。他们认为华盛顿关于出口管制的讨论被行业劫持,让中国公司获得了太多先进的芯片制造设计和机械。政府官员引用了商务部与为芯片行业工作并游说反对出口管制的律师事务所之间的旋转门,尽管这些官员也是政府中少数了解半导体供应链复杂性的人之一。因为这旋转门,特朗普政府官员认为,法规允许太多的技术泄露,削弱了美国相对于中国的地位。

The China hawks on the National Security Council concluded that America’s semiconductor industry needed to be saved from itself. Left to the whim of their shareholders and to market forces, chip firms would slowly transfer staff, technology, and intellectual property to China until Silicon Valley was hollowed out. The U.S. needed a stronger export control regime, the China hawks believed. They thought Washington’s discussion about export controls had been hijacked by the industry, letting Chinese firms acquire too much advanced chipmaking design and machinery. Administration officials cited the revolving door between the Commerce Department and law firms who worked for the chip industry and lobbied against export controls, though these officials were also among the few people in the government who understood the complexity of semiconductor supply chains. Because of this revolving door, Trump administration officials believed, regulations allowed too much technological leakage, weakening America’s position relative to China.

在特朗普总统的推特推文中,大多数人几乎没有注意到政府的不同部门——从国会到商务部,从白宫到五角大楼——正在以华盛顿自 1980 年代后期以来从未见过的方式重新关注半导体. 媒体关注焦点集中在特朗普与北京的“贸易战”和他的关税上调,精心宣布以最大限度地提高媒体关注度。特朗普征收关税的众多产品中包括芯片,导致一些分析师将半导体视为主要是贸易问题。然而,在政府的国家安全官僚机构中,总统的关税和他的贸易战被视为分散了人们对正在进行的高风险技术斗争的注意力。

Amid the fire and fury of President Trump’s Twitter feed, most people barely noticed how different parts of the government—from Congress to the Commerce Department, from the White House to the Pentagon—were refocusing on semiconductors in ways unseen in Washington since the late 1980s. Media attention focused on Trump’s “trade war” with Beijing and his tariff hikes, carefully announced to maximize media attention. Among the many products that Trump imposed tariffs on were chips, causing some analysts to see semiconductors as mostly a trade issue. Within the government’s national security bureaucracy, though, the president’s tariffs and his trade war were seen as a distraction from the high-stakes technological struggle underway.

2018 年 4 月,随着特朗普与中国的贸易争端升级,美国政府得出结论,中兴通讯向美国官员提供虚假信息违反了认罪协议的条款。据一位助手称,特朗普的商务部长威尔伯罗斯“非常个人化”,因为他在去年与中兴通讯的谈判中发挥了作用。商务部开始重新限制美国公司向中兴通讯出售产品的能力,这一决定通过了官僚机构“几乎没有人知道,”一位参与者说。当规则突然恢复时,中兴通讯再次被切断了购买美国半导体等产品的能力。如果美国不改变政策,该公司将走向倒闭。

In April 2018, as Trump’s trade dispute with China escalated, the U.S. government concluded that ZTE had violated the terms of its plea agreement by providing false information to U.S. officials. Wilbur Ross, Trump’s commerce secretary, took it “very personally,” according to one aide, since he’d played a role in negotiating the deal with ZTE the previous year. The Commerce Department began reimposing the restrictions on U.S. firms’ ability to sell to ZTE, a decision that moved through the bureaucracy “almost without anyone knowing,” according to one participant. When the rules snapped back, ZTE was again cut off from its ability to buy U.S. semiconductors, among other products. If the U.S. didn’t change policy, the company would careen toward collapse.

然而,特朗普本人对贸易比对技术更感兴趣。他将中兴通讯的潜在扼杀视为对习近平的筹码。因此,当中国领导人提议达成协议时,特朗普急切地接受了这个提议,并在推特上表示,他会想办法让中兴通讯继续经营,而不是担心中兴通讯。“在中国失去太多工作。” 很快,中兴通讯同意支付另一笔罚款,以换取重新获得美国供应商的准入。特朗普认为他在贸易战中获得了筹码,尽管这被证明是虚幻的。华盛顿的中国鹰派认为他被财政部长史蒂文·姆努钦(Steven Mnuchin)等官员欺骗了,后者一再敦促特朗普向北京做出让步。中兴的传奇首先展示了全球所有主要科技公司对美国芯片的依赖程度。正如一位政府官员所说,半导体不仅仅是“我们竞争的一切”的“基石”。它们也可能是一种毁灭性的强大武器。

Trump himself was more interested in trade than technology, however. He saw the potential strangulation of ZTE simply as leverage over Xi Jinping. So when the Chinese leader proposed doing a deal, Trump eagerly accepted the offer, tweeting that he’d find a way to keep ZTE in business out of concern for the company “losing too many jobs in China.” Soon ZTE agreed to pay another fine in exchange for regaining access to U.S. suppliers. Trump thought he’d gained leverage in the trade war, though this proved illusory. Washington’s China hawks thought he’d been duped by officials like Treasury Secretary Steven Mnuchin, who repeatedly urged Trump to offer concessions to Beijing. What the ZTE saga showed above all was the extent to which all the world’s major tech firms relied on U.S. chips. Semiconductors weren’t simply the “cornerstone” of “everything we’re competing on,” as one administration official had put it. They could also be a devastatingly powerful weapon.

第 50 章 福建金华

CHAPTER 50 Fujian Jinhua

清除计算机数据,” Kenny Wang 输入谷歌,搜索一个程序来掩盖他从美光网络下载机密文件时的踪迹。对 Google 的结果不满意,他尝试了不同的搜索。“清除计算机使用记录,”他输入。最终,他找到并运行了一个名为 CCleaner 的程序,显然是想从公司提供的惠普笔记本电脑上擦除文件。这并没有阻止调查人员发现他从他的雇主、美国存储芯片冠军美光公司那里下载了 900 个文件,并将这些文件放在 USB 驱动器上并上传到 Google Drive。“Micron Confidential / Do Not Duplicate”,这些文件被贴上了标签。王不是简单地复制文件:他计划复制美光制造尖端 DRAM 芯片的秘方,下载详细说明美光芯片布局的文件,该公司如何为光刻工艺制造掩模的详细信息,以及测试和产量的详细信息——这些秘密将美光估计,这需要几年时间和数亿美元来复制。

“Clear computer data,” Kenny Wang typed into Google, searching for a program to cover his tracks as he downloaded confidential files from Micron’s network. Unsatisfied with Google’s results, he tried a different search. “Clear computer use records,” he entered. Eventually he found and ran a program called CCleaner, apparently trying to wipe files off his company-supplied HP laptop. This didn’t stop investigators from discovering he’d downloaded nine hundred files from his employer, Micron, America’s memory chip champion, which he put on a USB drive and uploaded to Google Drive. “Micron Confidential / Do Not Duplicate,” the files were labeled. Wang wasn’t simply duplicating files: he planned to duplicate Micron’s secret recipe for cutting-edge DRAM chips, downloading files detailing Micron’s chip layouts, details for how the company made masks for its lithography processes, and test and yield details—secrets that would have taken several years and hundreds of millions of dollars to replicate, Micron estimated.

今天,三家公司主导着全球 DRAM 芯片市场,美光及其两家韩国竞争对手三星和 SK 海力士。台湾公司在 1990 年代和 2000 年代花费数十亿美元试图打入 DRAM 业务,但从未成功建立盈利业务。DRAM 市场需要规模经济,因此小型生产商很难具有价格竞争力。虽然台湾日本和韩国在 1970 年代和 1980 年代首次进入芯片行业时都专注于 DRAM 芯片。DRAM 需要专业知识、先进设备和大量资本投资。先进的设备通常可以从美国、日本和荷兰的大型工具制造商处购买现成的。诀窍是困难的部分。当三星在 1980 年代后期进入该行业时,它从美光公司获得技术许可,在硅谷开设了一家研发机构,并聘请了数十名受过美国培训的博士。另一种更快的获取技术诀窍的方法是挖走员工并窃取文件。

Three companies dominate the world’s market for DRAM chips today, Micron and its two Korean rivals, Samsung and SK Hynix. Taiwanese firms spent billions trying to break into the DRAM business in the 1990s and 2000s but never managed to establish profitable businesses. The DRAM market requires economies of scale, so it’s difficult for small producers to be price competitive. Though Taiwan never succeeded in building a sustainable memory chip industry, both Japan and South Korea had focused on DRAM chips when they first entered the chip industry in the 1970s and 1980s. DRAM requires specialized know-how, advanced equipment, and large quantities of capital investment. Advanced equipment can generally be purchased off-the-shelf from the big American, Japanese, and Dutch toolmakers. The know-how is the hard part. When Samsung entered the business in the late 1980s, it licensed technology from Micron, opened an R&D facility in Silicon Valley, and hired dozens of American-trained PhDs. Another, faster, method for acquiring know-how is to poach employees and steal files.

中国福建省与台湾隔海相望。福建历史悠久的港口城市厦门的港口坐落着台湾控制的金门岛,毛泽东的军队在冷战最紧张的时刻多次炮击该岛。台湾与福建省的关系密切,但并不总是友好的。然而,当福建省政府决定开设一家名为金华的 DRAM 芯片制造商并为其提供超过 50 亿美元的政府资助,金华认为与台湾合作是其成功的最佳途径。台湾没有任何领先的存储芯片公司,但它确实拥有美光在 2013 年购买的 DRAM 设施。

China’s Fujian Province is right across the straits from Taiwan. In the harbor of Fujian’s historic port city of Xiamen sits the Taiwanese-controlled island of Kinmen, which Mao Zedong’s armies repeatedly shelled during the tensest moments of the Cold War. The relationship between Taiwan and Fujian Province is close but not always friendly. Yet when the government of Fujian Province decided to open a DRAM chipmaker called Jinhua and provided it with over $5 billion in government funding, Jinhua wagered that a partnership with Taiwan was its best path to success. Taiwan didn’t have any leading memory chip companies, but it did have DRAM facilities, which Micron had purchased in 2013.

美光不会向金华提供任何帮助,它认为金华是一个危险的竞争对手。如果晋华能够学会掌握 DRAM 技术,那么它获得的巨额政府补贴将提供主要的竞争优势,让它以廉价芯片充斥 DRAM 市场,从而降低美光、三星和海力士的利润率。三大 DRAM 公司花了几十年的时间投资于超专业化的技术工艺,这不仅创造了地球上最先进的存储芯片,而且还定期进行改进和降低成本。他们的专业知识受到专利的保护,但更重要的是只有他们的工程师拥有的专业知识。

Micron wasn’t going to provide any help to Jinhua, which it saw as a dangerous competitor. If Jinhua could ever learn to master DRAM technology, the massive government subsidies it received would provide a major competitive advantage, letting it flood the DRAM market with cheap chips, reducing profit margins at Micron, Samsung, and Hynix. The big three DRAM firms had spent decades investing in ultra-specialized technology processes, which not only created the most advanced memory chips on earth, but also had produced a regular cadence of improvements and cost reductions. Their expertise was defended by patents, but even more important was the know-how that only their engineers had.

为了竞争,金华必须以公平或不公平的方式获得这种制造技术。芯片行业收购竞争对手技术的历史由来已久,可以追溯到 1980 年代有关日本知识产权盗窃的一系列指控。然而,金华的技术更接近于克格勃的 T 局。首先,金华与台湾联电达成了一项协议,后者制造逻辑芯片(不是存储芯片),联电将获得约 7 亿美元,以换取提供生产 DRAM 的专业知识。许可协议在半导体行业很常见,但该协议有一个转折点。UMC 承诺提供 DRAM 技术,但它不在DRAM业务中。因此,在 2015 年 9 月,联电从美光台湾工厂聘请了多名员工,首先是总裁史蒂文·陈,他负责开发联电的 DRAM 技术并管理其与金华的关系。下个月,UMC 在美光的台湾工厂聘请了一位名叫 JT Ho 的流程经理。在接下来的一年里,何鸿燊收到了他的前美光同事 Kenny Wang 的一系列文件,他仍在爱达荷州芯片制造商的台湾工厂工作。最终,Wang 离开了 Micron,来到了 UMC,带来了与他一起上传到 Google Drive 的九百个文件。

To compete, Jinhua had to acquire this manufacturing know-how by means fair or foul. There’s a long history in the chip industry of acquiring rivals’ technology, dating back to the string of allegations about Japanese intellectual property theft in the 1980s. Jinhua’s technique, however, was closer to the KGB’s Directorate T. First, Jinhua cut a deal with Taiwan’s UMC, which fabricated logic chips (not memory chips), whereby UMC would receive around $700 million in exchange for providing expertise in producing DRAM. Licensing agreements are common in the semiconductor industry, but this agreement had a twist. UMC was promising to provide DRAM technology, but it wasn’t in the DRAM business. So in September 2015, UMC hired multiple employees from Micron’s facility in Taiwan, starting with the president, Steven Chen, who was put in charge of developing UMC’s DRAM technology and managing its relationship with Jinhua. The next month, UMC hired a process manager at Micron’s Taiwan facility named J. T. Ho. Over the subsequent year, Ho received a series of documents from his former Micron colleague, Kenny Wang, who was still working at the Idaho chipmaker’s facility in Taiwan. Eventually, Wang left Micron to move to UMC, bringing nine hundred files uploaded to Google Drive with him.

美光通知台湾检方有关阴谋,并开始收集证据窃听王的手机。他们很快就积累了足够的证据来对联电提出指控,联电后来就从美光窃取的一些技术申请了专利。当美光起诉联电和金华侵犯其专利时,他们在中国福建省进行了反诉。福建一家法院裁定,美光应对侵犯联电和金华的专利负责——这些专利是使用从美光窃取的材料提交的。福州市中级人民法院为“补救”禁止美光在该公司最大的市场中国销售 26 种产品。

Taiwanese prosecutors were notified by Micron of the conspiracy and started gathering evidence by tapping Wang’s phone. They soon accumulated enough evidence to bring charges against UMC, which had since filed for patents on some of the technology it stole from Micron. When Micron sued UMC and Jinhua for violating its patents, they countersued in China’s Fujian Province. A Fujian court ruled that Micron was responsible for violating UMC and Jinhua’s patents—patents that had been filed using material stolen from Micron. To “remedy” the situation, Fuzhou Intermediate People’s Court banned Micron from selling twenty-six products in China, the company’s biggest market.

这是一个完美的案例研究国家支持的知识产权盗窃,在中国经营的外国公司长期以来一直抱怨的。当然,台湾人自然明白为什么中国人宁愿不遵守知识产权规则。德州仪器在 1960 年代首次抵达台湾时,李国栋部长曾嘲讽“知识产权是帝国主义欺负落后国家。” 然而,台湾已经得出结论,最好尊重知识产权规范,尤其是在其公司开始开发自己的技术并拥有自己的专利要捍卫的情况下。许多知识产权专家预测,随着中国公司生产更复杂的产品,中国将很快开始窃取更少的知识产权。然而,这篇论文的证据好坏参半。奥巴马政府与中国间谍机构达成协议,同意停止向中国公司提供被盗机密的努力持续了很长时间,以至于美国人忘记了这个问题。及时重启。

This was a perfect case study of the state-backed intellectual property theft foreign companies operating in China had long complained of. The Taiwanese naturally understood why the Chinese preferred not to abide by intellectual property rules, of course. When Texas Instruments first arrived in Taiwan in the 1960s, Minister K. T. Li had sneered that “intellectual property rights are how imperialists bully backward countries.” Yet Taiwan had concluded it was better to respect intellectual property norms, especially as its companies began developing their own technologies and had their own patents to defend. Many intellectual property experts predicted that China would soon begin stealing less IP as its companies produced more sophisticated goods. However, the evidence for this thesis was mixed. Efforts by the Obama administration to cut a deal with China’s spy agencies whereby they agreed to stop providing stolen secrets to Chinese companies lasted only long enough for Americans to forget about the issue, at which point the hacking promptly restarted.

美光没有理由期待在中国得到公平审判。当福建的袋鼠法庭可以将公司排除在其最大的市场之外时,在台湾或加利福尼亚打赢官司意义不大。大约在同一时间,美国半导体制造设备生产商 Veeco 在美国法院对中国竞争对手 AMEC 提起了知识产权诉讼,该诉讼在福建省法院提起反诉,该省法院是美光的竞争对手所在的省份。一名纽约法官发布了有利于 Veeco 的初步禁令。根据伯克利大学教授、中国法律专家马克·科恩的研究,福建法院以自己的初步禁令进行报复,禁止 Veeco 向中国进口机械设备,这一举措仅发生在 0.01% 的中国专利案件中。鉴于美国的法庭案件耗时数月,福建法院仅用了九个工作日就作出了判决。裁决本身是仍然是秘密。

Micron had little reason to expect a fair trial in China. Winning court cases in Taiwan or California meant little when kangaroo courts in Fujian could lock the company out of its biggest market. Around the same time, Veeco, an American producer of semiconductor manufacturing equipment, had launched an intellectual property suit in U.S. courts against a Chinese competitor, AMEC, which countersued in a Fujian provincial court—the same province where Micron’s competitor was located. A New York judge issued a preliminary injunction in Veeco’s favor. The Fujian court retaliated with a preliminary injunction of its own, banning Veeco from importing machinery to China, a move that occurs in only 0.01 percent of Chinese patent cases, according to research by Berkeley professor Mark Cohen, an expert on Chinese law. Whereas the U.S. court case took months, the Fujian court reached its decision in just nine business days. The ruling itself is still secret.

美光似乎将面临类似的命运。由于美光的秘密掌握在金华手中,一些分析人士认为,金华在几年后就可以大规模生产 DRAM 芯片——到那时,美光是否能重新进入中国市场并不重要,因为金华将生产芯片使用美光科技并以补贴价格出售。如果这发生在奥巴马政府期间,此案将导致严厉的声明,但没有别的。美国的首席执行官们知道他们不能指望美国政府的大力支持,他们会试图与北京达成协议,放弃他们的知识产权,以期重新进入中国市场。金华知道没有什么比愤怒的新闻稿更糟糕的了,他会尽可能地压榨公司。其他外国公司会保持沉默,即使他们知道他们可能是下一个。

Micron seemed set to face a similar fate. With Micron’s secrets at Jinhua’s disposal, some analysts thought it would only be a few years before Jinhua was producing DRAM chips at scale—at which point it wouldn’t matter if Micron was let back into the Chinese market, because Jinhua would be producing chips using Micron’s technology and selling them at subsidized prices. Had this occurred during the Obama administration, the case would have resulted in stern statements but little else. American CEOs, knowing they couldn’t count on serious U.S. government backing, would have tried to cut a deal with Beijing, surrendering their intellectual property in hopes of regaining access to the Chinese market. Jinhua, knowing to expect nothing worse than an angry press release, would have squeezed the company as hard as it could. Other foreign firms would have stayed quiet even though they knew they could be next.

NSC 中的对华鹰派决心改变这种局面。他们将美光案视为特朗普承诺要解决的不公平贸易类型,尽管总统本人对美光没有表现出特别的兴趣。一些政府官员主张利用奥巴马总统 2015 年签署的网络间谍行政命令中规定的权力对金华实施金融制裁,尽管该命令尚未用于针对中国一家大型公司。深思熟虑后,特朗普政府决定使用与中兴通讯相同的工具,理由是通过贸易监管来解决贸易争端更有意义。金华被禁止购买美国制造芯片的设备。

The China hawks on the NSC were determined to change this dynamic. They saw the Micron case as the type of unfair trade that Trump had promised to fix, even though the president himself displayed no particular interest in Micron. Some administration officials advocated imposing financial sanctions on Jinhua, using powers set out in an executive order on cyber espionage signed by President Obama in 2015, though the order hadn’t been used against a major Chinese company. After deliberating, the Trump administration decided to use the same tool it had deployed against ZTE, reasoning that it made more sense to address a trade dispute with a trade regulation. Jinhua was cut off from buying U.S. equipment for manufacturing chips.

应用材料公司、Lam Research 和 KLA 等美国公司是生产不可替代机械的少数寡头公司的一部分,例如在硅晶片上沉积微观薄层材料或识别纳米级缺陷的工具。如果没有这种机器——其中大部分仍然在美国制造——就不可能生产出先进的半导体。只有日本有公司生产一些类似的机器,所以如果东京和华盛顿同意,他们可以让任何国家的任何公司都无法制造先进的芯片。在与强大的日本经济、贸易和工业部的官员进行详细磋商后,特朗普政府相信东京支持对晋华采取强硬举措,并将确保日本公司不会削弱美国对该公司的限制。这给了美国使世界任何地方的任何芯片制造商都倒闭的工具。特朗普政府中的一些鸽派,比如财政部长姆努钦,都很紧张。但有权实施出口管制的商务部长威尔伯罗斯认为“我们他妈的为什么不用这个?” 据一位助手说。因此,在金华向提供其关键芯片制造工具的美国公司支付发票后,美国禁止了他们的出口。几个月内,在金华生产地面停了下来。中国最先进的DRAM公司被摧毁。

U.S. companies like Applied Materials, Lam Research, and KLA are part of a small oligopoly of companies that produce irreplaceable machinery, like the tools that deposit microscopically thin layers of materials on silicon wafers or recognize nanometer-scale defects. Without this machinery—much of it still built in the U.S.—it’s impossible to produce advanced semiconductors. Only Japan has companies producing some comparable machinery, so if Tokyo and Washington agreed, they could make it impossible for any firm, in any country, to make advanced chips. After detailed consultations with officials at Japan’s powerful Ministry of Economics, Trade, and Industry, the Trump administration was confident Tokyo supported a tough move against Jinhua and would ensure Japanese companies didn’t undercut American restrictions on the firm. This gave the U.S. a powerful new tool to put out of business any chipmaker, anywhere in the world. Some of the doves in the Trump administration, like Treasury Secretary Mnuchin, were nervous. But Commerce Secretary Wilbur Ross, who had the authority to impose export controls, thought “why the fuck wouldn’t we use this?” according to one aide. So after Jinhua paid invoices to the U.S. firms that supplied its crucial chipmaking tools, the U.S. banned their export. Within months, production at Jinhua ground to a halt. China’s most advanced DRAM firm was destroyed.

第 51 章 对华为的攻击

CHAPTER 51 The Assault on Huawei

当被问及华为时,特朗普总统向他最喜欢的电视节目之一Fox & Friends的主持人解释说: 称之为间谍通道“我们不希望他们的设备在美国,因为他们在监视我们……。他们什么都知道。” 技术基础设施可用于窃取机密信息几乎不是什么启示。2013 年,前国家安全局雇员爱德华·斯诺登在泄露该机构许多最机密的秘密后叛逃到俄罗斯后,世界各地的报纸定期讨论美国网络侦探能力的消息。中国令人印象深刻的黑客能力在一系列高调泄露表面上是秘密的美国政府数据之后也广为人知。

“I call it the spyway,” President Trump explained to the hosts of Fox & Friends, one of his favorite TV programs, when asked about Huawei. “We don’t want their equipment in the United States because they spy on us…. They know everything.” It was hardly a revelation that tech infrastructure could be used to pilfer confidential information. After former National Security Agency employee Edward Snowden defected to Russia in 2013 while releasing many of the agency’s most closely held secrets, news of American cyber sleuths’ capabilities were regularly discussed in the world’s newspapers. China’s impressive hacking capabilities were also well known after a string of high-profile breaches of ostensibly secret U.S. government data.

在五角大楼和国家安全委员会内部,华为与其说是一场间谍挑战——尽管美国官员毫不怀疑该公司会支持中国的间谍活动——不如说是长期争夺技术主导地位的第一场战斗。五角大楼官员马特·特平(Matt Turpin)曾研究过军方的新抵消战略,他认为华为是美国科技行业更广泛问题的征兆:中国公司“实际上与美国同在系统内”,因为他们设计了芯片使用美国软件,使用美国机器生产它们,并经常将它们插入为美国消费者制造的设备中。鉴于此,“美国不可能‘创新’中国,然后否认他们的创新成果。” 华为和其他中国公司在美国认为需要主导的科技子行业发挥核心作用,以保持对中国的技术优势,无论是军事上还是战略上。“华为真的变成了代表我们在与中国的科技竞争中做错的一切,”另一位特朗普政府高级官员表示。

Within the Pentagon and the NSC, Huawei was seen less as an espionage challenge—though U.S. officials had little doubt the company would support Chinese spycraft—than as the first battle in a long struggle for technological dominance. Matt Turpin, a Pentagon official who’d worked on the military’s new offset strategy, saw Huawei as symptomatic of a broader problem in the U.S. tech industry: Chinese firms “were effectively inside the system with the United States,” given that they designed chips with U.S. software, produced them using U.S. machinery, and often plugged them into devices built for American consumers. Given this, it was impossible “for the United States to ‘out-innovate’ China and then deny them the fruits of that innovation.” Huawei and other Chinese firms were assuming central roles in tech subsectors that the U.S. thought it needed to dominate to retain a technological advantage over China, militarily and strategically. “Huawei became really a proxy for everything we had done wrong with our tech competition with China,” another senior Trump administration official put it.

对华为的担忧不仅限于特朗普政府或美国。澳大利亚在其安全服务部门认定风险根本无法减轻后,已禁止华为进入 5G 网络,即使华为交出了对其所有软件源代码和硬件的访问权限。澳大利亚总理马尔科姆·特恩布尔起初对彻底禁令持怀疑态度。据澳大利亚记者 Peter Hartcher 称,特恩布尔给自己买了一本 474 页的书《5G 安全综合指南》来研究这个主题,这样他就可以向他的技术专家提出更好的问题。最终,他确信他别无选择,只能取缔该公司。澳大利亚成为第一个正式将华为设备从其 5G 网络中移除的国家,日本、新西兰等国很快就做出了这一决定。

Concern about Huawei wasn’t confined to the Trump administration or the United States. Australia had banned Huawei from 5G networks after its security services concluded the risk simply couldn’t be mitigated, even if Huawei turned over access to all its software source code and hardware. Australian prime minister Malcolm Turnbull had at first been skeptical of an outright ban. According to Australian journalist Peter Hartcher, Turnbull bought himself a 474-page-book titled A Comprehensive Guide to 5G Security to study the topic so that he could ask better questions of his tech experts. Eventually he was convinced he had no choice but to ban the firm. Australia became the first country to formally cut Huawei’s equipment from its 5G networks, a decision that was soon followed by Japan, New Zealand, and others.

并非每个国家都有相同的威胁评估。许多中国邻国对该公司持怀疑态度,不愿承担网络安全风险。相比之下,在欧洲,几个传统的美国盟友对特朗普政府说服他们禁止华为的施压行动持谨慎态度。美国在东欧的一些亲密盟友公开禁止该公司,例如波兰,该公司也在 2019 年以间谍罪名逮捕了一名前公司高管。法国还有悄悄地施加了严格的限制。其他欧洲大国试图找到中间立场。向中国出口大量汽车和机械的德国受到中国驻华大使的警告如果它禁止华为,“后果”。“中国政府不会袖手旁观,”这位中国外交官威胁说。

Not every country had the same threat assessment. Many of China’s neighbors were skeptical of the company and unwilling to take risks with network security. In Europe, by contrast, several traditional American allies looked warily at the Trump administration’s pressure campaign to convince them to ban Huawei. Some close American allies in Eastern Europe openly banned the company, like Poland, which also in 2019 arrested a former company executive on espionage charges. France also quietly imposed strict restrictions. Other big European countries tried to find a middle ground. Germany, which exports large quantities of cars and machinery to China, was warned by the Chinese ambassador of “consequences” if it banned Huawei. “The Chinese government will not stand idly by,” the Chinese diplomat threatened.

最终,特朗普政府预计会遭到德国的抵制,德国将其视为在一系列问题上搭便车的盟友。更大的惊喜是英国,尽管它与美国拒绝美国要求禁止华为进入英国的 5G 网络,而是从瑞典的爱立信或芬兰的诺基亚等替代供应商那里购买设备。2019 年,英国政府的国家网络安全中心得出结论,华为系统的风险可以在没有禁令的情况下进行管理。

Ultimately the Trump administration expected pushback from Germany, which it saw as a free-riding ally on a range of issues. The bigger surprise was Britain, which despite its “special relationship” with the United States was spurning U.S. requests to ban Huawei from the UK’s 5G networks and, instead, buy equipment from alternative suppliers like Sweden’s Ericsson or Finland’s Nokia. In 2019, the UK government’s National Cyber Security Centre concluded the risk of Huawei systems could be managed without a ban.

为什么澳大利亚和英国的网络安全专家对华为风险的评估存在差异?没有证据表明存在技术分歧。英国监管机构对例如,华为网络安全实践中的缺陷。辩论实际上是关于是否应该阻止中国在世界科技基础设施中发挥越来越大的作用。英国信号情报机构前负责人罗伯特·汉尼根认为,“我们应该接受中国未来将成为全球科技强国,现在就开始管理风险,而不是假装西方可以坐视中国的科技崛起。” 许多欧洲人还认为中国的技术进步是不可避免的,因此不值得阻止。

Why did Australian and British cybersecurity experts differ in their assessment of Huawei risk? There’s no evidence of technical disagreements. UK regulators were quite critical of deficiencies in Huawei’s cybersecurity practices, for example. The debate was really about whether China should be stopped from playing an ever-larger role in the world’s tech infrastructure. Robert Hannigan, former head of the UK’s signals intelligence agency, argued that “we should accept that China will be a global tech power in the future and start managing the risk now, rather than pretending the west can sit out China’s technological rise.” Many Europeans also thought China’s technological advance was inevitable and therefore not worth trying to stop.

美国政府不同意。华为的问题远远超出了关于该公司是否帮助窃听电话或窃取数据的争论。华为高管承认他们会违反美国对伊朗的制裁激怒了华盛顿的许多人,但最终只是一场戏。真正的问题是,中华人民共和国的一家公司已经迈上了技术阶梯——从 1980 年代后期的简单电话交换机到 2010 年代后期最先进的电信和网络设备。它的年度研发支出现在可以与微软、谷歌和英特尔等美国科技巨头相媲美。在所有中国科技公司中,它是最成功的出口商,它对国外市场有详细的了解。它不仅为手机信号塔生产硬件,还设计了尖端的智能手机芯片。它已成为台积电的第二大客户,仅次于苹果。紧迫的问题是:美国能让这样的中国公司成功吗?

The United States government didn’t agree. The issue with Huawei went far beyond the debate over whether the company helped tap phones or pilfer data. Huawei executives’ admission that they’d violated U.S. sanctions on Iran angered many in Washington but was ultimately a sideshow. The real issue was that a company in the People’s Republic of China had marched up the technology ladder—from, in the late 1980s, simple phone switches to, by the late 2010s, the most advanced telecom and networking gear. Its annual R&D spending now rivaled American tech giants like Microsoft, Google, and Intel. Of all China’s tech firms, it was the most successful exporter, giving it detailed knowledge of foreign markets. It not only produced hardware for cell towers, it also designed cutting-edge smartphone chips. It had become TSMC’s second biggest customer, behind only Apple. The pressing question was: Could the United States let a Chinese company like this succeed?

像这样的问题让华盛顿的许多人感到不舒服。一代人以来,美国的精英们欢迎并促成了中国的经济崛起。美国也鼓励技术亚洲的公司,在日本快速增长的几年里为索尼等日本公司提供市场准入,并在几十年后为韩国三星提供市场准入。华为的商业模式与索尼或三星首次在世界科技生态系统中获得重要地位时的商业模式并没有太大区别。多一点竞争不是好事吗?

Questions like this made many people in Washington uncomfortable. For a generation, America’s elite had welcomed and enabled China’s economic rise. The United States had also encouraged technology companies across Asia, providing market access to Japanese firms like Sony during the years of Japan’s rapid growth and doing the same for South Korea’s Samsung several decades later. Huawei’s business model wasn’t much different from that of Sony or Samsung when they first won a major position in the world’s tech ecosystem. Wasn’t a bit more competition a good thing?

然而,在国家安全委员会上,与中国的竞争现在主要是零和游戏。这些官员将华为解读为不是商业挑战,而是战略挑战。索尼和三星是总部设在与美国结盟国家的科技公司,华为是美国主要地缘政治对手的全国冠军。从这个角度来看,华为的扩张是一种威胁。国会也想要一个更严厉、更具战斗性的政策。“美国需要扼杀华为,”共和党参议员本·萨斯 (Ben Sasse) 在 2020 年宣称。“现代战争是用半导体打的,我们让华为使用我们的美国设计。”

On the National Security Council, however, competition with China was now seen primarily in zero-sum terms. These officials interpreted Huawei not as a commercial challenge but as a strategic one. Sony and Samsung were tech firms based in countries that were allied with the U.S. Huawei was a national champion of America’s primary geopolitical rival. Viewed through this lens, Huawei’s expansion was a threat. Congress wanted a tougher, more combative policy, too. “The United States needs to strangle Huawei,” Republican senator Ben Sasse declared in 2020. “Modern wars are fought with semiconductors and we were letting Huawei use our American designs.”

重点不在于华为直接支持中国军队,而在于该公司正在提升中国芯片设计和微电子技术的整体水平。这个国家生产的电子产品越先进,它购买的尖端芯片就越多,世界半导体生态系统就越依赖中国,而以美国为代价。此外,针对中国最知名的科技公司将在全球范围内发出信息,警告其他国家准备选边站队。阻碍华为的崛起成为政府的执念。

The point was less that Huawei was directly supporting China’s military than that the company was advancing China’s overall level of chip design and microelectronics know-how. The more advanced electronics the country produced, the more cutting-edge chips it would buy, and the more the world’s semiconductor ecosystem would rely on China, at the expense of the United States. Moreover, targeting China’s highest-profile tech firm would send a message worldwide, warning other countries to prepare to take sides. Hobbling Huawei’s rise became a fixation of the administration.

当特朗普政府第一次决定对华为施加压力时,它禁止向该公司出售美国制造的芯片。鉴于英特尔芯片无处不在,而且许多其他美国公司生产几乎不可替代的模拟芯片,这一限制本身就是毁灭性的。然而,经过数十年的离岸外包,在美国进行的半导体生产过程比以前少得多。例如,华为生产的芯片不是在美国设计的——美国缺乏能够构建先进设备的设施智能手机处理器——但在台湾台积电。限制向华为出口美国制造的商品不会阻止台积电为华为制造先进芯片。

When the Trump administration first decided to turn up its pressure on Huawei, it prohibited the sale of U.S.-made chips to the company. This restriction alone was devastating, given that Intel chips are ubiquitous and many other U.S. companies manufacture all-but-irreplaceable analog chips. Yet after decades of offshoring, far less of the semiconductor production process took place in the United States than previously. For example, Huawei produced the chips that it designed not in the U.S.—which lacked facilities capable of building advanced smartphone processors—but at Taiwan’s TSMC. Restricting the export of U.S.-made goods to Huawei would do nothing to stop TSMC from fabricating advanced chips for Huawei.

人们可能预计芯片制造的离岸外包会降低美国政府限制先进芯片制造的能力。如果世界上所有先进的芯片制造仍然在美国的土地上,那么切断华为肯定会更容易。然而,美国仍有牌可打。例如,离岸芯片制造过程恰逢芯片行业瓶颈日益垄断。世界上几乎所有的芯片都使用至少来自三个美国公司之一的软件:Cadence、Synopsys 和 Mentor(后者归德国西门子所有,但总部位于俄勒冈州)。除了英特尔内部制造的芯片外,所有最先进的逻辑芯片都是由三星和台积电这两家公司制造的,这两家公司都位于依赖美国军方保障安全的国家。此外,制造先进的处理器需要由荷兰 ASML 一家公司生产的 EUV 光刻机,而后者又依赖其圣地亚哥子公司 Cymer(它于 2013 年收购)为其 EUV 光刻工具提供不可替代的光源. 当许多基本步骤需要由少数几家公司生产的工具、材料或软件时,控制芯片制造过程中的瓶颈要容易得多。许多这些瓶颈仍然在美国人手中。那些不是主要由美国亲密盟友控制的。当许多基本步骤需要由少数几家公司生产的工具、材料或软件时,控制芯片制造过程中的瓶颈要容易得多。许多这些瓶颈仍然在美国人手中。那些不是主要由美国亲密盟友控制的。当许多基本步骤需要由少数几家公司生产的工具、材料或软件时,控制芯片制造过程中的瓶颈要容易得多。许多这些瓶颈仍然在美国人手中。那些不是主要由美国亲密盟友控制的。

One might have expected the offshoring of chipmaking to have reduced the U.S. government’s ability to restrict access to advanced chip fabrication. It would certainly have been easier to cut off Huawei if all the world’s advanced chipmaking was still based on U.S. soil. However, the U.S. still had cards to play. For example, the process of offshoring chip fabrication had coincided with a growing monopolization of chip industry choke points. Nearly every chip in the world uses software from at least one of three U.S.-based companies, Cadence, Synopsys, and Mentor (the latter of which is owned by Germany’s Siemens but based in Oregon). Excluding the chips Intel builds in-house, all the most advanced logic chips are fabricated by just two companies, Samsung and TSMC, both located in countries that rely on the U.S. military for their security. Moreover, making advanced processors requires EUV lithography machines produced by just one company, the Netherlands’ ASML, which in turn relies on its San Diego subsidiary, Cymer (which it purchased in 2013), to supply the irreplaceable light sources in its EUV lithography tools. It’s far easier to control choke points in the chipmaking process when so many essential steps require tools, materials, or software produced by just a handful of firms. Many of these choke points remained in American hands. Those that didn’t were mostly controlled by close U.S. allies.

大约在这个时候,两位学者亨利·法瑞尔和亚伯拉罕·纽曼注意到国际政治和经济关系越来越受到他们所谓的“武器化的相互依存。” 他们指出,各国比以往任何时候都更加紧密相连,但相互依存不是化解冲突和鼓励合作,而是创造了新的竞争场所。将国家联系在一起的网络已成为冲突的领域。例如,在金融领域,美国将其他国家依赖银行系统的行为作为武器来惩罚伊朗。这些学者担心美国政府将贸易和资本流动作为政治武器威胁全球化,并冒着危险的意外后果的风险。相比之下,特朗普政府得出结论,它拥有将半导体供应链武器化的独特权力。

Around this time, two academics, Henry Farrell and Abraham Newman, noticed that international political and economic relations were increasingly impacted by what they called “weaponized interdependence.” Countries were more intwined than ever, they pointed out, but rather than defusing conflicts and encouraging cooperation, interdependence was creating new venues for competition. Networks that knit together nations had become a domain of conflict. In the financial sphere, the U.S. had weaponized other countries’ reliance on access to the banking system to punish Iran, for example. These academics worried that the U.S. government’s use of trade and capital flows as political weapons threatened globalization and risked dangerous unintended consequences. The Trump administration, by contrast, concluded it had unique power to weaponize semiconductor supply chains.

2020 年 5 月,政府进一步收紧对华为的限制。现在,商务部宣布,它将“通过限制华为使用美国技术和软件在国外设计和制造半导体的能力来保护美国国家安全”。商务部的新规定并没有简单地停止向华为出售美国生产的商品。他们也限制任何使用美国生产技术制造的商品出售给华为。在充满瓶颈的芯片行业中,这几乎意味着任何芯片。不使用美国制造设备,台积电无法为华为制造先进芯片。没有美国生产的软件,华为就无法设计芯片。即使是中国最先进的代工厂中芯国际,也广泛依赖美国工具。华为只是与世界整个芯片制造基础设施隔绝,除了美国的芯片

In May 2020, the administration tightened restrictions on Huawei further. Now, the Commerce Department declared, it would “protect U.S. national security by restricting Huawei’s ability to use U.S. technology and software to design and manufacture its semiconductors abroad.” The new Commerce Department rules didn’t simply stop the sale of U.S.-produced goods to Huawei. They restricted any goods made with U.S.-produced technology from being sold to Huawei, too. In a chip industry full of choke points, this meant almost any chip. TSMC can’t fabricate advanced chips for Huawei without using U.S. manufacturing equipment. Huawei can’t design chips without U.S.-produced software. Even China’s most advanced foundry, SMIC, relies extensively on U.S. tools. Huawei was simply cut off from the world’s entire chipmaking infrastructure, except for chips that the U.S. Commerce Department deigned to give it a special license to buy.

世界芯片业迅速开始执行美国的规则。尽管美国试图剔除其第二大客户,但台积电董事长刘强东承诺不仅要遵守法律条文,而且也是它的精神。“这不仅可以通过对规则的解释来解决,而且与美国政府的意图有关,”他告诉记者。自那时候起,华为被迫剥离部分智能手机业务和服务器业务无法获得必要的筹码。中国推出自己的 5G 电信网络,这曾是政府备受瞩目的优先事项由于芯片短缺而推迟。在美国实施限制后,其他国家,尤其是英国,决定禁止华为,理由是在没有美国芯片的情况下,该公司将难以为其产品提供服务。

The world’s chip industry quickly began implementing the U.S. rules. Even though the U.S. was trying to eviscerate its second-largest customer, TSMC’s chairman, Mark Liu, promised not only to abide by the letter of the law but also its spirit. “This is something that can be solved not solely through the interpretation of the rules, but also has to do with the intentions of the U.S. government,” he told journalists. Since then, Huawei’s been forced to divest part of its smartphone business and its server business, since it can’t get the necessary chips. China’s rollout of its own 5G telecoms network, which was once a high-profile government priority, has been delayed due to chip shortages. After the U.S. restrictions took place, other countries, notably Britain, decided to ban Huawei, reasoning that in the absence of U.S. chips the company would struggle to service its products.

在对华为的攻击之后,将其他多家中国科技公司列入黑名单。与美国商量后,荷兰决定不批准出售ASML 的 EUV 机器供应给中国公司。2017 年被 AMD 描述为“战略合作伙伴”的超级计算机公司曙光在 2019 年被美国列入黑名单。据《华盛顿邮报》报道,美国官员称 Phytium 公司为用于测试高超音速导弹的超级计算机设计了芯片。Phytium 的芯片是使用美国软件设计的,并由台积电在台湾生产。进入美国及其盟友的半导体生态系统促进了 Phytium 的发展。然而,该公司对外国软件和制造的依赖使其极易受到美国限制的影响。

The assault on Huawei was followed by blacklisting multiple other Chinese tech firms. After discussions with the United States, the Netherlands decided not to approve the sale of ASML’s EUV machines to Chinese firms. Sugon, the supercomputer company that AMD described in 2017 as a “strategic partner,” was blacklisted by the U.S. in 2019. So, too, was Phytium, a company that U.S. officials say has designed chips for supercomputers that were used to test hypersonic missiles, according to a report in the Washington Post. Phytium’s chips were designed using U.S. software and produced in Taiwan at TSMC. Access to the semiconductor ecosystem of America and its allies enabled Phytium’s growth. However, the company’s reliance on foreign software and manufacturing left it critically vulnerable to U.S. restrictions.

不过,归根结底,美国对中国科技公司的攻击只是一次有限的打击。许多中国最大的科技公司,如腾讯和阿里巴巴,在购买美国芯片或让台积电制造半导体的能力方面仍然没有具体限制。中芯国际是中国最先进的逻辑芯片生产商,在购买先进芯片制造工具方面面临新的限制,但并未停业。甚至华为也被允许购买较旧的半导体,例如用于连接 4G 网络的半导体。

Ultimately, though, the American assault on China’s tech firms has been a limited strike. Many of China’s biggest tech companies, like Tencent and Alibaba, still face no specific limits on their purchases of U.S. chips or their ability to have TSMC manufacture their semiconductors. SMIC, China’s most advanced producer of logic chips, faces new restrictions on its purchases of advanced chipmaking tools, but it has not been put out of business. Even Huawei is allowed to buy older semiconductors, like those used for connecting to 4G networks.

然而,令人惊讶的是,中国没有采取任何行动来报复其最具全球性的科技公司的步履蹒跚。它一再威胁要惩罚美国科技公司,但从未扣动扳机。北京表示正在制定危害中国安全的外国公司的“不可靠实体清单”,但似乎没有将任何公司添加到清单中。北京显然已经盘算,与其对美国进行反击,不如接受华为将成为二流技术玩家。事实证明,在切断供应链方面,美国在升级方面具有主导地位。“武器化的相互依存,”一位前高级官员在华为遭到袭击后沉思道。“它的美好的事物。”

Nevertheless, it’s surprising that China’s done nothing to retaliate against the hobbling of its most global tech firm. It has repeatedly threatened to punish U.S. tech firms but never pulled the trigger. Beijing said it was drawing up an “unreliable entity list” of foreign companies that endanger Chinese security, but it doesn’t appear to have added any firms to the list. Beijing has evidently calculated that it’s better to accept that Huawei will become a second-rate technology player than to hit back against the United States. The U.S., it turns out, has escalation dominance when it comes to severing supply chains. “Weaponized interdependence,” one former senior official mused after the strike on Huawei. “It’s a beautiful thing.”

第 52 章 中国的人造卫星时刻?

CHAPTER 52 China’s Sputnik Moment?

W当中国城市武汉于 2020 年 1 月 23 日在 COVID-19 病例海啸中被封锁时,它在大流行的任何时候都面临着任何城市中最严厉、最长的限制。COVID病毒及其引起的疾病仍然知之甚少。中国政府一直压制对该病毒的讨论,直到它席卷武汉并在中国和世界范围内传播。政府姗姗来迟地关闭了进出武汉的旅行,在城市周边设置了检查站,关闭了企业,并命令该市几乎所有的 1000 万人在封锁结束前不要离开他们的公寓。以前从未有过如此庞大的大都市被冻结。高速公路空无一人,人行道荒凉,机场和火车站关闭。除了医院和杂货店,几乎所有地方都关门了。

When the Chinese city of Wuhan locked down on January 23, 2020, amid a tsunami of cases of COVID-19, it faced some of the harshest, longest restrictions of any city at any point in the pandemic. The COVID virus and the disease it caused was still little understood. China’s government had suppressed discussion of the virus until it ripped through Wuhan and was spreading across China and the world. The government belatedly shut down travel in and out of Wuhan, imposing checkpoints on the city’s perimeter, shuttering businesses, and ordering almost all the city’s 10 million people not to leave their apartments until the lockdown ended. Never before had such a massive metropolis simply frozen. Highways were empty, sidewalks desolate, airports and train stations closed. Except for hospitals and grocery stores, almost everything was shut.

除了一个设施,那就是。总部位于武汉的扬子存储器技术公司 (YMTC) 是中国领先的 NAND 存储器生产商,NAND 存储器是一种广泛存在于从智能手机到 USB 记忆棒的消费设备中的芯片。今天有五家公司生产具有竞争力的 NAND 芯片;没有一家总部设在中国。然而,许多行业专家认为,在所有类型的芯片中,中国实现世界级制造能力的最佳机会在于 NAND 生产。清华紫光,半导体巨头投资全球芯片公司的基金,与中国国家芯片基金和省政府一道,为长江存储提供了至少 240 亿美元的资金。

Except for one facility, that is. Yangzte Memory Technologies Corporation (YMTC), based in Wuhan, is China’s leading producer of NAND memory, a type of chip that’s ubiquitous in consumer devices from smartphones to USB memory sticks. There are five companies that make competitive NAND chips today; none are headquartered in China. Many industry experts, however, think that of all types of chips, China’s best chance at achieving world-class manufacturing capabilities is in NAND production. Tsinghua Unigroup, the semiconductor slush fund that invested in chip companies worldwide, provided YMTC with at least $24 billion in funding, alongside China’s national chip fund and the provincial government.

据日本报纸《日经亚洲》报道,中国政府对长江存储的支持是如此之大,以至于即使在 COVID 封锁期间,它也被允许继续工作。途经武汉的列车搭载长江存储员工专用客车,让他们在封城期间进入武汉。该公司甚至在 2020 年 2 月下旬和 3 月上旬招聘武汉的职位,由于该国其他地区仍处于冻结状态。中国领导人几乎愿意为抗击冠状病毒做任何事情,但他们优先考虑建立半导体产业。

So great is China’s government support for YMTC that even during the COVID lockdown it was allowed to keep working, according to Nikkei Asia, a Japanese newspaper with some of the best coverage of China’s chip industry. Trains passing through Wuhan carried special passenger cars specifically for YMTC employees, letting them enter Wuhan despite the lockdown. The company was even hiring for Wuhan-based positions in late February and early March 2020, as the rest of the country remained frozen. China’s leaders were willing to do almost anything in their fight against the coronavirus, but their effort to build a semiconductor industry took priority.

人们普遍认为,与美国不断升级的科技竞争对中国政府来说就像是“人造卫星时刻”。暗指美国在 1957 年人造卫星发射后担心它落后于竞争对手,促使华盛顿向科学和技术投入资金。在美国禁止向华为等公司销售芯片后,中国无疑面临着人造卫星规模的冲击。王丹是中国科技政策最聪明的分析师之一,他认为美国的限制已经通过推动支持芯片行业的新政府政策,“推动了北京对技术主导地位的追求”。他认为,如果没有美国新的出口管制,《中国制造 2025》最终会像中国之前的产业政策努力一样,政府会浪费大量资金。由于美国的压力,中国政府可能会为中国芯片制造商提供比他们原本无法获得的更多支持。

It’s commonly argued that the escalating tech competition with the United States is like a “Sputnik moment” for China’s government. The allusion is to the United States’ fear after the launch of Sputnik in 1957 that it was falling behind its rival, driving Washington to pour funding into science and technology. China certainly faced a Sputnik-scale shock after the U.S. banned sales of chips to firms like Huawei. Dan Wang, one of the smartest analysts of China’s tech policy, has argued that American restrictions have “boosted Beijing’s quest for tech dominance” by catalyzing new government policies to support the chip industry. In the absence of America’s new export controls, he argues, Made in China 2025 would have ended up like China’s previous industrial policy efforts, with the government wasting substantial sums of money. Thanks to U.S. pressure, China’s government may provide Chinese chipmakers more support than they’d otherwise have received.

争论的焦点是美国是否应该试图破坏中国不断发展的芯片生态系统——从而引发不可避免的反击——或者在希望中国芯片驱动逐渐消失的同时在国内投资是否更明智。美国的限制措施无疑催生了政府对中国芯片制造商的新一波支持。习近平最近任命他的最高经济助手刘鹤为“芯片沙皇”管理该国的半导体工作。毫无疑问,中国花费数十亿美元补贴芯片公司。这笔资金是否会产生新技术还有待观察。例如,武汉市不仅是中国对 NAND 芯片平价的最大希望的长江存储的所在地,也是该国近期最大的半导体骗局的所在地。

The debate is about whether the U.S. should try to derail China’s growing chip ecosystem—thereby spurring an inevitable counterreaction—or whether it’s smarter simply to invest at home while hoping China’s chip drive peters out. U.S. restrictions have certainly catalyzed a new wave of government support for Chinese chipmakers. Xi Jinping recently appointed his top economic aide, Liu He, to serve as a “chip czar,” managing the country’s semiconductor efforts. There’s no doubt that China’s spending billions to subsidize chip firms. Whether this funding produces new technology remains to be seen. For example, the city of Wuhan is home not only to YMTC, China’s brightest hope for NAND chip parity, but also to the country’s biggest recent semiconductor scam.

武汉宏芯(HSMC)的案例显示了在没有提出足够问题的情况下将资金投入半导体的风险。根据已从互联网上删除的中国媒体报道,HSMC 是由一群骗子创立的,他们持有写有“台积电-副总裁”的假名片,并散布谣言说他们的亲戚是共产党的高级官员。他们欺骗武汉当地政府投资他们的公司,然后用这笔资金聘请台积电前研发负责人担任首席执行官。随着他的加入,他们从 ASML 购买了一台深紫外光刻机,然后利用这一壮举从投资者那里筹集了更多资金。但是武汉的工厂是台积电旧工厂的粗制滥造的复制品。公司倒闭时,HSMC 仍在尝试生产其第一款芯片。

The case of Wuhan Hongxin (HSMC) shows the risk of shoveling money into semiconductors without asking enough questions. According to a Chinese media report that’s since been removed from the internet, HSMC was founded by a group of scam artists who carried fake business cards that read “TSMC—Vice President” and spread rumors that their relatives were top Communist Party officials. They duped the Wuhan local government into investing in their company, then used the funds to hire as CEO TSMC’s former head of R&D. With him on board, they acquired a deep-ultraviolet lithography machine from ASML, then used this feat to raise more funds from investors. But the factory in Wuhan was a shoddily built copy of an old TSMC facility; HSMC was still trying to produce its first chip when the company went bust.

失败的不仅仅是省级实验。清华紫光集团最近在其全球收购狂潮之后用光了现金,并拖欠了部分债券。即便是清华 CEO 赵卫国的高层政治关系也不足以拯救这家公司,尽管它所拥有的芯片公司很可能会安然无恙地生存下去。中国政府规划部门的一位官员公开感叹中国的芯片产业“没有经验,没有技术,没有天赋。” 这是夸大其词,但很明显,数十亿美元已在中国浪费在半导体项目上,这些项目要么完全不切实际,要么像 HSMC 那样公然欺诈。如果中国的人造卫星时刻激发了更多类似这样的国家支持的半导体项目,那么该国将不会走上技术独立的道路。

It isn’t only provincial experiments that have failed. Tsinghua Unigroup recently ran out of cash after its global acquisition spree and defaulted on some of its bonds. Even Tsinghua CEO Zhao Weiguo’s top-level political connections weren’t enough to save the firm, though the chip companies it owns will likely survive mostly unscathed. An official from China’s government planning agency publicly lamented that the country’s chip industry had “no experience, no technology, no talent.” This is an overstatement, but it’s clear that billions of dollars have been wasted in China on semiconductor projects that are either hopelessly unrealistic or, like HSMC, blatant frauds. If China’s Sputnik moment inspires more state-backed semiconductor programs like these, the country won’t be on a path to technological independence.

在拥有如此跨国供应链的行业中,技术独立始终是一个白日梦,即使对于仍然是世界上最大的半导体参与者的美国来说也是如此。中国在供应链的许多环节都缺乏有竞争力的公司,从机械到软件,技术独立更是难上加难。为了完全独立,中国需要获得尖端的设计软件、设计能力、先进材料和制造技术,以及其他步骤。中国无疑会在其中一些领域取得进展,但有些领域的成本太高,中国难以在国内复制。

In an industry with such a multinational supply chain, technological independence was always a pipe dream, even for the United States, which remains the world’s biggest semiconductor player. For China, which lacks competitive firms in many parts of the supply chain, from machinery to software, technological independence is even more difficult. For complete independence, China would need to acquire cutting-edge design software, design capabilities, advanced materials, and fabrication know-how, among other steps. China will no doubt make progress in some of these spheres, yet some are simply too expensive and too difficult for China to replicate at home.

例如,考虑复制 ASML 的一台 EUV 机器需要什么,这台机器花了近 30 年的时间开发和商业化。EUV 机器具有多个组件,这些组件本身就构成了极其复杂的工程挑战。仅复制 EUV 系统中的激光器就需要完美识别和组装 457,329 个零件。单个缺陷可能会导致严重的延迟或可靠性问题。毫无疑问,中国政府已经部署了一些最优秀的间谍来研究 ASML 的生产过程。然而,即使他们已经侵入了相关系统并下载了设计规范,这种复杂的机器也不能像被盗文件一样简单地复制和粘贴。即使间谍要获得专业信息,他们也需要光学或激光博士学位才​​能理解科学——即使如此,

Consider, for example, what it would take to replicate one of ASML’s EUV machines, which have taken nearly three decades to develop and commercialize. EUV machines have multiple components that, on their own, constitute epically complex engineering challenges. Replicating just the laser in an EUV system requires perfectly identifying and assembling 457,329 parts. A single defect could cause debilitating delays or reliability problems. No doubt the Chinese government has deployed some of its best spies to study ASML’s production processes. However, even if they’ve already hacked into the relevant systems and downloaded design specs, machinery this complex can’t simply be copied and pasted like a stolen file. Even if a spy were to gain access to specialized information, they’d need a PhD in optics or lasers to understand the science—and even still, they’d lack the three decades of experience accumulated by the engineers who’ve developed EUV.

或许在十年内,中国可以成功建造自己的 EUV 扫描仪。如果是这样,该计划将耗资数百亿美元,但是——这必然会令人沮丧——一旦准备就绪,它将不再是最前沿的。届时,ASML 将推出新一代工具,称为高孔径 EUV,计划于 2020 年代中期准备就绪,每台机器的成本为 3 亿美元,是第一代 EUV 机器成本的两倍。即使未来的中国 EUV 扫描仪和 ASML 目前的设备一样好——很难想象,因为美国将试图限制其从其他国家获取组件的能力——使用这种假设性替代 EUV 机器的中国芯片制造商也将难以盈利有了它,因为到 2030 年,台积电、三星和英特尔将已经使用自己的 EUV 扫描仪十年,在此期间,他们将完善他们的使用并支付这些工具的成本。他们将能够以远低于中国公司使用假想的中国制造的 EUV 工具的价格出售使用 EUV 生产的芯片。

Perhaps in a decade China can succeed in building its own EUV scanner. If so, the program will cost tens of billions of dollars, but—in a revelation that is bound to be discouraging—when it’s ready it will no longer be cutting edge. By that time, ASML will have introduced a new generation tool, called high-aperture EUV, which is scheduled to be ready in the mid-2020s and cost $300 million per machine, twice the cost of the first generation EUV machine. Even if a future Chinese EUV scanner works just as well as ASML’s current equipment—hard to imagine, given that the U.S. will try to restrict its ability to access components from other countries—Chinese chipmakers using this hypothetical alternative EUV machine will struggle to produce profitably with it, because by 2030, TSMC, Samsung, and Intel will have already used their own EUV scanners for a decade, during which time, they’ll have perfected their use and paid down the cost of these tools. They’ll be able to sell chips produced with EUV for far cheaper than a Chinese company using a hypothetical Chinese-built EUV tool.

EUV 机器只是通过跨国供应链生产的众多工具之一。将供应链的每一个环节都驯化起来,成本高得不可思议。全球芯片行业每年的资本支出超过 1000 亿美元。除了建立目前缺乏的专业知识和设施基础外,中国还必须复制这一支出。建立一个尖端的全国内供应链将需要十多年的时间,并且在此期间花费超过一万亿美元。

EUV machines are just one of many tools that are produced via multinational supply chains. Domesticating every part of the supply chain would be impossibly expensive. The global chip industry spends over $100 billion annually on capital expenditures. China would have to replicate this spending in addition to building a base of expertise and facilities that it currently lacks. Establishing a cutting-edge, all-domestic supply chain would take over a decade and cost well over a trillion dollars in that period.

这就是为什么尽管有言辞,但中国实际上并没有追求全国内供应链。北京承认这根本不可能。中国希望建立非美国的供应链,但由于美国在芯片行业的影响力和其出口法规的域外权力,非美国的供应链也是不现实的,除非是在遥远的将来。中国在某些领域减少对美国的依赖并增加其在芯片行业的整体权重,尽可能多地摆脱瓶颈技术,这似乎是合理的。

This is why, despite the rhetoric, China’s not actually pursuing an all-domestic supply chain. Beijing recognizes this is simply impossible. China would like a non-U.S. supply chain, but because of America’s heft in the chip industry and the extraterritorial power of its export regulations, a non-American supply chain is also unrealistic, except perhaps in the distant future. What is plausible is for China to reduce its reliance on the United States in certain spheres and to increase its overall weight in the chip industry, weaning itself off as many choke-point technologies as possible.

中国今天的核心挑战之一是许多芯片要么使用 x86 架构(用于 PC 和服务器),要么使用 Arm 架构(用于移动设备);x86 由两家美国公司 Intel 和 AMD 主导,而授权其他公司使用其架构的 Arm 则位于英国。但是,现在有一种名为 RISC-V 的新指令集架构是开源的,因此任何人都可以免费使用它。开源架构的想法吸引了芯片行业的许多部门。目前必须向 Arm 支付许可证的任何人都希望获得免费的替代方案。此外,安全缺陷的风险可能会更低,因为像 RISC-V 这样的开源架构的开放性意味着更多的工程师将能够验证细节并识别错误。出于同样的原因,创新的步伐也可能更快。与开发 RISC-V 相关的各种项目。中国公司也接受了 RISC-V,因为他们认为它在地缘政治上是中立的。2019 年,管理架构的 RISC-V 基金会,出于这个原因,他从美国搬到了瑞士。考虑到这一点,阿里巴巴等公司正在设计基于 RISC-V 架构的处理器。

One of China’s core challenges today is that many chips use either the x86 architecture (for PCs and servers) or the Arm architecture (for mobile devices); x86 is dominated by two U.S. firms, Intel and AMD, while Arm, which licenses other companies to use its architecture, is based in the UK. However, there’s now a new instruction set architecture called RISC-V that is open-sourced, so it’s available to anyone without a fee. The idea of an open-source architecture appeals to many parts of the chip industry. Anyone who currently must pay Arm for a license would prefer a free alternative. Moreover, the risk of security defects may be lower, because the open nature of an open-source architecture like RISC-V means that more engineers will be able to verify details and identify errors. For the same reason, the pace of innovation may be faster, too. These two factors explain why DARPA has funded a variety of projects related to developing RISC-V. Chinese firms have also embraced RISC-V, because they see it as geopolitically neutral. In 2019, the RISC-V Foundation, which manages the architecture, moved from the U.S. to Switzerland for this reason. Companies like Alibaba are designing processors based on the RISC-V architecture with this in mind.

除了与新兴架构合作外,中国还专注于使用较旧的工艺技术来构建逻辑芯片。智能手机和数据中心需要最尖端的芯片,但汽车和其他消费设备通常使用较旧的工艺技术,这种技术足够强大且便宜得多。中国新晶圆厂的大部分投资,包括对中芯国际等公司的投资,都集中在落后节点的产能上。中芯国际已经表明,中国拥有生产具有竞争力的落后逻辑芯片的劳动力。即使美国的出口限制越来越严格,他们也不太可能禁止出口已有数十年历史的制造设备。中国也在大力投资新兴的半导体材料,如碳化硅和氮化镓,这不太可能取代大多数芯片中的纯硅,但可能会在管理电动汽车的电力系统方面发挥更大的作用。在这里,中国可能也有必要的技术,所以政府补贴可能有助于它在价格上赢得业务。

In addition to working with emerging architectures, China’s also focusing on older process technology to build logic chips. Smartphones and data centers require the most cutting-edge chips, but cars and other consumer devices often use older process technology, which is sufficiently powerful and far cheaper. Most of the investment in new fabs in China, including at companies like SMIC, is in production capacity at lagging-edge nodes. SMIC has already shown that China has the workforce to produce competitive lagging-edge logic chips. Even if U.S. export restrictions get tighter, they’re unlikely to prohibit the export of decades-old manufacturing equipment. China’s also investing heavily in emerging semiconductor materials like silicon carbide and gallium nitride, which are unlikely to displace pure silicon in most chips but will likely play a bigger role in managing the power systems in electric vehicles. Here, too, China probably has the requisite technology, so government subsidies may help it win business on price.

其他国家担心的是,中国的大量补贴将让它在供应链的多个部分赢得市场份额,尤其是那些不需要最先进技术的部分。除非对获得外国软件和机器实施严格的新限制,否则中国似乎可能在生产非尖端逻辑芯片方面发挥更大的作用。此外,它还在为开发电动汽车电源管理芯片所需的材料投入大量资金。与此同时,中国的长江存储真正有机会赢得 NAND 内存市场的一大块。在整个芯片行业,估计表明,到 2030 年,中国的制造份额将从本世纪初的 15% 增加到全球产能的 24%,在数量上超过台湾和韩国。几乎可以肯定,中国在技术上仍会落后。但如果更多的芯片产业转移到中国,该国将在要求技术转让方面拥有更大的影响力。美国和其他国家实施出口限制的成本将变得更高,而中国将拥有更广泛的劳动力来源。中国几乎所有的芯片公司都依赖于政府的支持,因此它们既面向国家目标,也面向商业目标。一位高管对《日经亚洲报》表示,“盈利和上市……不是长江存储的首要任务” 。相反,该公司专注于“打造国家自己的芯片和实现中国梦。”

The worry for other countries is that China’s slew of subsidies will let it win market share across multiple parts of the supply chain, especially those that don’t require the most advanced technologies. Barring severe new restrictions on access to foreign software and machinery, China looks likely to play a much bigger role in producing non-cutting-edge logic chips. In addition, it’s pouring money into the materials needed to develop power management chips for electric vehicles. China’s YMTC, meanwhile, has a real chance to win a chunk of the NAND memory market. Across the chip industry, estimates suggest that China’s share of fabrication will increase from 15 percent at the start of the decade to 24 percent of global capacity by 2030, overtaking Taiwan and South Korea in terms of volume. China will almost certainly still lag technologically. But if more of the chip industry moves to China, the country will have more leverage in demanding technology transfer. It will become more costly for the U.S. and other countries to impose export restrictions, and China will have a broader pool of workers from which to draw. Almost all of China’s chip firms are dependent on government support, so they’re oriented toward national goals as much as commercial ones. “Making profits and going public… are not the priority” at YMTC, one executive told the Nikkei Asia newspaper. Instead, the company’s focused on “building the country’s own chips and realizing the Chinese dream.”

第 53 章 短缺和供应链

CHAPTER 53 Shortages and Supply Chains

作为一个国家,或者太久了,我们还没有进行我们需要的大而大胆的投资,以超越我们的全球竞争对手,”拜登总统对一众首席执行官宣布。坐在白宫里,泰迪·罗斯福的画作下,高高举着​​一块 12 英寸的硅片,拜登看着 Zoom 屏幕,斥责高管们“在研发和制造方面落后......我们必须加强我们的比赛,”他告诉他们。屏幕上的十九位高管中的许多人都同意了。为了讨论美国对芯片短缺的反应,拜登邀请了台积电等外国公司和英特尔等美国芯片制造商,以及遭受严重半导体短缺的半导体知名用户。福特和通用汽车的 CEO 通常不会被邀请参加有关芯片的高层会议,通常他们也不会感兴趣。但在 2021 年期间,随着世界经济及其供应链在大流行引发的中断之间摇摆不定,世界各地的人们开始了解他们的生活以及他们的生计在多大程度上依赖于半导体。

“For too long as a nation, we haven’t been making the big, bold investments we need to outpace our global competitors,” President Biden declared to a screenful of CEOs. Sitting in the White House under a painting of Teddy Roosevelt, holding aloft a twelve-inch silicon wafer, Biden looked into the Zoom screen and castigated the executives for “falling behind on research and development and manufacturing…. We have to step up our game,” he told them. Many of the nineteen executives on the screen agreed. To discuss America’s response to the chip shortage, Biden invited foreign companies like TSMC alongside U.S. chipmakers like Intel, as well as prominent users of semiconductors who were suffering severe semiconductor shortages. The CEOs of Ford and GM weren’t normally invited to high-level meetings about chips, and normally they wouldn’t have been interested. But over the course of 2021, as the world’s economy and its supply chains convulsed between pandemic-induced disruptions, people around the world began to understand just how much their lives, and often their livelihoods, depended on semiconductors.

2020 年,就在美国开始对中国实施芯片扼流圈,切断该国一些领先的科技公司获得美国芯片技术的机会时,第二次芯片扼流圈开始扼杀世界经济的一部分。某些类型的芯片变得难以尤其是广泛应用于汽车的基础逻辑芯片种类。这两个芯片扼流圈部分相关。至少自 2019 年以来,华为等中国公司就一直在囤积芯片,为美国未来可能的制裁做准备,而中国晶圆厂则尽可能多地购买制造设备,以防美国决定收紧对芯片制造工具的出口限制。

In 2020, just as the United States began to impose a chip choke on China, cutting off some of the country’s leading tech companies from accessing U.S. chip technology, a second chip choke began asphyxiating parts of the world economy. Certain types of chips became difficult to acquire, especially the types of basic logic chips that are widely used in automobiles. The two chip chokes were partially interrelated. Chinese firms like Huawei had been stockpiling chips since at least 2019, in preparation for potential future U.S. sanctions, while Chinese fabs were buying as much manufacturing equipment as possible in case the U.S. decided to tighten export restrictions on chipmaking tools.

然而,中国的库存仅解释了 COVID 时代芯片扼流圈的部分原因。更大的原因是大流行开始后芯片订单大幅波动,因为公司和消费者调整了对不同商品的需求。2020 年,随着数百万人升级电脑以在家工作,PC 需求激增。随着越来越多的生活转向在线,数据中心对服务器的需求也在增长。汽车公司最初削减芯片订单,预计汽车销量将下滑。当需求迅速恢复时,他们发现芯片制造商已经将产能重新分配给其他客户。根据行业组织美国汽车政策委员会的数据,世界上最大的汽车公司可以在每辆汽车中使用超过一千个芯片。如果缺少一个芯片,汽车就无法发货。汽车制造商在 2021 年的大部分时间里都在苦苦挣扎,而且往往未能收购半导体。据估计,这些公司在 2021 年生产的汽车数量比如果它们可能生产的汽车数量减少了 770 万辆没有面临芯片短缺,这意味着根据行业估计,集体收入损失为 2100 亿美元。

However, Chinese stockpiling explains only part of the COVID-era chip choke. The bigger cause is vast swings in orders for chips after the pandemic began, as companies and consumers adjusted their demand for different goods. PC demand spiked in 2020, as millions of people upgraded their computers to work from home. Data centers’ demand for servers grew, too, as more of life shifted online. Car companies at first cut chip orders, expecting car sales to slump. When demand quickly recovered, they found that chipmakers had already reallocated capacity to other customers. According to the American Automotive Policy Council, an industry group, the world’s biggest auto companies can use over a thousand chips in each car. If even one chip is missing, the car can’t be shipped. Carmakers spent much of 2021 struggling and often failing to acquire semiconductors. These firms are estimated to have produced 7.7 million fewer cars in 2021 than would have been possible had they not faced chip shortages, which implies a $210 billion collective revenue loss, according to industry estimates.

拜登政府和大多数媒体将芯片短缺解释为供应链问题。白宫委托撰写了一份长达 250 页的关于半导体供应链漏洞的报告。然而,半导体短缺主要不是由芯片供应链问题引起的。出现了一些供应中断,例如马来西亚的 COVID 封锁,这影响了那里的半导体封装业务。但研究公司 IC Insights 的数据显示,2021 年全球生产的芯片比以往任何时候都多——超过 1.1 万亿个半导体设备。这是一个与 2020 年相比增长 13%。半导体短缺主要是需求增长而非供应问题。它是由新的个人电脑、5G 手机、支持人工智能的数据中心——以及最终我们对计算能力的永不满足的需求。

The Biden administration and most of the media interpreted the chip shortage as a supply chain problem. The White House commissioned a 250-page report on supply chain vulnerabilities that focused on semiconductors. However, the semiconductor shortage wasn’t primarily caused by issues in the chip supply chain. There were some supply disruptions, like COVID lockdowns in Malaysia, which impacted semiconductor packaging operations there. But the world produced more chips in 2021 than ever before—over 1.1 trillion semiconductor devices, according to research firm IC Insights. This was a 13 percent increase compared to 2020. The semiconductor shortage is mostly a story of demand growth rather than supply issues. It’s driven by new PCs, 5G phones, AI-enabled data centers—and, ultimately, our insatiable demand for computing power.

世界各地的政客因此误诊了半导体供应链的困境。问题不在于芯片行业广泛的生产过程对 COVID 和由此产生的封锁处理不佳。很少有行业在这场大流行中受到如此小的干扰。出现的此类问题,尤其是汽车芯片短缺,主要是由于汽车制造商在大流行初期疯狂且不明智地取消了芯片订单,再加上他们的准时制制造实践几乎没有提供足够的利润。错误。对于收入遭受数千亿美元打击的汽车行业来说,有充分的理由重新思考他们如何管理自己的供应链。然而,半导体行业迎来了辉煌的一年。除了一场大地震(一种低但非零概率的风险)之外,很难想象和平时期对供应链的冲击比该行业自 2020 年初以来幸存下来的冲击更严重。2020 年和 2021 年芯片产量的大幅增长并不是跨国供应链断裂的迹象。这是他们工作的迹象。

Politicians around the world have therefore misdiagnosed the semiconductor supply chain dilemma. The problem isn’t that the chip industry’s far-flung production processes dealt poorly with COVID and the resulting lockdowns. There are few industries that sailed through the pandemic with so little disruption. Such problems that emerged, notably the shortage of auto chips, are mostly the fault of carmakers’ frantic and ill-advised cancelation of chip orders in the early days of the pandemic coupled with their just-in-time manufacturing practices that provide little margin of error. For the car industry, which suffered a several-hundred-billion-dollar hit to revenue, there’s plenty of reason to rethink how they’ve managed their own supply chains. The semiconductor industry, however, had a banner year. Besides a massive earthquake—a low but non-zero probability risk—it’s hard to imagine a more severe peacetime shock to supply chains than what the industry has survived since early 2020. The substantial increase in chip production during both 2020 and 2021 is not a sign that multinational supply chains are broken. It’s a sign that they’ve worked.

尽管如此,政府应该比过去更加认真地考虑半导体供应链。过去几年真正的供应链教训不是关于脆弱性,而是关于利润和权力。台湾的非凡崛起表明,一家拥有远见和政府财政支持的公司如何能够重塑整个行业。与此同时,美国对中国获得芯片技术的限制表明,芯片行业的瓶颈有多么强大。然而,过去十年中国半导体产业的崛起提醒我们,这些瓶颈并非无限持久。国家和政府通常可以找到绕过阻塞点的方法,尽管这样做既费时又昂贵,有时甚至非常昂贵。技术转变也会削弱阻塞点的功效。

Nevertheless, governments should think harder about semiconductor supply chains than they used to. The real supply chain lesson of the past few years is not about fragility but about profits and power. Taiwan’s extraordinary ascent shows how one company—with a vision and with government financial support—can remake an entire industry. Meanwhile, U.S. restrictions on China’s access to chip technology demonstrate just how powerful the chip industry’s choke points are. The rise of China’s semiconductor industry over the past decade, however, is a reminder that these choke points are not infinitely durable. Countries and governments can often find ways around choke points, though doing so is time-consuming and expensive, sometimes extraordinarily so. Technological shifts can erode the efficacy of choke points, too.

这些阻塞点只有在由几家公司控制时才有效,理想情况下只有一家公司控制。尽管拜登政府已答应工作“与行业、盟友和合作伙伴”,在谈到芯片行业的未来时,美国及其盟友并不完全一致。美国希望扭转其芯片制造份额下降的趋势,并保持其在半导体设计和机械领域的主导地位。而欧洲和亚洲国家则希望在高价值芯片设计市场中获得更大的份额。与此同时,台湾和韩国没有计划放弃制造先进逻辑和存储芯片的市场领先地位。由于中国将扩大自己的制造能力视为国家安全的必要条件,未来可以在美国、欧洲和亚洲之间共享的芯片制造业务数量有限。美国要增加市场份额,就必须减少其他国家的市场份额。美国 隐含地希望从拥有现代芯片制造设施的其他领域之一抢占市场份额。然而在中国之外,世界上所有先进的芯片厂都在美国的盟友或密友国家。

These choke points only work if they’re controlled by a couple of companies, and ideally only by one. Although the Biden administration has promised to work “with industry, allies, and partners,” the U.S. and its allies aren’t completely aligned when it comes to the future of the chip industry. The U.S. wants to reverse its declining share of chip fabrication and retain its dominant position in semiconductor design and machinery. Countries in Europe and Asia, however, would like to grab a bigger share of the high-value chip design market. Taiwan and South Korea, meanwhile, have no plans to surrender their market-leading positions fabricating advanced logic and memory chips. With China viewing expansion of its own fabrication capacity as a national security necessity, there’s a limited amount of future chip fabrication business that can be shared between the U.S., Europe, and Asia. If the U.S. wants to increase its market share, some other country’s market share must decrease. The U.S. is implicitly hoping to grab market share from one of the other areas with modern chipmaking facilities. Yet outside China, all the world’s advanced chip fabs are in countries that are U.S. allies or close friends.

然而,韩国计划保持其在制造存储芯片方面的领先地位,同时试图扩大其在制造逻辑芯片方面的作用。韩国总统文在寅指出:“半导体企业之间的竞争现在已经开始在各国蔓延。” “我的政府也将作为一个团队与企业合作,让韩国保持半导体强国的地位。” 韩国政府向平泽市投入了大量资金,该市以前是美国军事基地的所在地,但现在是三星一家主要工厂的所在地。从应用材料公司到东京电子,所有主要的芯片制造设备公司都在该市开设了办事处。三星表示,除了在存储芯片生产上投入相当的资金外,它还计划到 2030 年在其逻辑芯片业务上投入超过 1000 亿美元。三星创始人李在镕的孙子于 2021 年被假释出狱,因受贿罪正在服刑。韩国法务省援引“经济因素”来证明他的释放是合理的,包括媒体报道暗示的,他将帮助公司做出重大半导体投资决策的预期。

South Korea, however, plans to retain its leading position in making memory chips while trying to expand its role in making logic chips. “Rivalries among semiconductor businesses have now begun to draw in countries,” South Korean president Moon Jae-in has noted. “My administration will also work with business as one team so Korea stays a semiconductor powerhouse.” The Korean government has poured money into a city called Pyeongtaek, formerly home to a U.S. military base but now the site of a major Samsung facility. All the major chipmaking equipment companies, from Applied Materials to Tokyo Electron, have opened offices in the city. Samsung has said it plans to spend over $100 billion by 2030 on its logic chip business in addition to investing comparable sums in memory chip production. The grandson of Samsung’s founder, Lee Jay-yong, was paroled from prison in 2021, where he was serving a sentence for bribery. Korea’s Justice Ministry cited “economic factors” in justifying his release, including, media reports suggested, expectations that he will help the company make major semiconductor investment decisions.

三星及其较小的韩国竞争对手 SK 海力士从韩国政府的支持中受益,但被困在中国和美国之间,每个国家都试图哄骗韩国的芯片巨头在本国建立更多的制造业。例如,三星最近宣布计划扩大和升级其在德克萨斯州奥斯汀生产先进逻辑芯片的设施,估计投资为 170 亿美元。然而,两家公司都面临着美国关于升级其在中国设施的提议的审查。美国施压据报道,限制将 EUV 工具转移到 SK 海力士在中国无锡的工厂,这正在推迟其现代化进程——并且可能会给公司带来大量成本。

Samsung and its smaller Korean rival SK Hynix benefit from the support of the Korean government but are stuck between China and the U.S., with each country trying to cajole South Korea’s chip giants to build more manufacturing in their countries. Samsung recently announced plans to expand and upgrade its facility for producing advanced logic chips in Austin, Texas, for example, an investment estimated to cost $17 billion. Both companies face scrutiny from the U.S. over proposals to upgrade their facilities in China, however. U.S. pressure to restrict the transfer of EUV tools to SK Hynix’s facility in Wuxi, China, is reportedly delaying its modernization—and presumably imposing a substantial cost on the company.

用文在寅总统的话来说,韩国并不是唯一一个芯片公司和政府作为“团队”开展工作的国家。台湾政府仍然大力保护其芯片产业,它认为这是其在国际舞台上最大的筹码来源。现在表面上从台积电完全退休的张忠谋已担任台湾贸易特使。他和台湾的主要兴趣仍然是确保台积电保持其在世界芯片行业的核心地位。该公司本身计划在 2022 年至 2024 年间投资超过 1000 亿美元,以升级其技术并扩大芯片制造能力。尽管该公司计划升级其在中国南京的工厂,并在亚利桑那州开设一家新工厂,但大部分资金将投资于台湾。然而,这些新晶圆厂都不会生产最先进的芯片,所以台积电最先进的技术将留在台湾。Chang继续呼吁半导体行业的“自由贸易”,并威胁说否则“成本将上升,技术发展将放缓”。与此同时,台湾政府多次介入支持台积电,通过维持台币贬值等措施支持台积电。使台湾出口更具竞争力。

South Korea isn’t the only country where chip companies and the government work as a “team,” to use President Moon’s phrase. Taiwan’s government remains fiercely protective of its chip industry, which it recognizes as its greatest source of leverage on the international stage. Morris Chang, now ostensibly fully retired from TSMC, has served as a trade envoy for Taiwan. His primary interest—and Taiwan’s—remains ensuring that TSMC retains its central role in the world’s chip industry. The company itself plans to invest over $100 billion between 2022 and 2024 to upgrade its technology and expand chipmaking capacity. Most of this money will be invested in Taiwan, though the company plans to upgrade its facility in Nanjing, China, and to open a new fab in Arizona. Neither of these new fabs will produce the most cutting-edge chips, however, so TSMC’s most advanced technology will remain in Taiwan. Chang continues to call for “free trade” in the semiconductor industry, threatening that otherwise “costs will go up, technology development will slow down.” Meanwhile, Taiwan’s government has repeatedly intervened to support TSMC through such measures as keeping Taiwan’s currency undervalued to make Taiwanese exports more competitive.

欧洲、日本和新加坡是另外三个寻求新半导体投资的地区。一些欧盟领导人建议欧洲大陆可以“大规模投资”并生产 3nm 或2nm 芯片,使欧洲晶圆厂接近尖端。鉴于非洲大陆在高级逻辑方面的市场份额较低,这不太可能。更有可能的是,欧洲将说服英特尔这样的大型外国芯片公司建造新工厂,为欧洲汽车制造商提供稳定的供应来源。新加坡继续为芯片制造提供大量激励措施,最近从美国 GlobalFoundries 获得了 40 亿美元的新晶圆厂投资。与此同时,日本正在大力补贴台积电建设新的芯片制造工厂与索尼合作。自从森田昭夫等高管离开日本后的几十年里,日本的芯片制造已经失去了很大一部分,但索尼仍然保留着一项规模可观且盈利的业务,​​制造可以感知图像并用于许多消费类设备的相机的半导体。不过,日本决定资助新台积电工厂的决定主要不是为了帮助索尼。日本政府担心,如果制造业继续向海外转移,日本保持强势地位的供应链部分,如机床和先进材料,也会转移到国外。

Europe, Japan, and Singapore are three other regions looking for new semiconductor investments. Some European Union leaders have suggested the continent can “invest massively” and produce 3nm or 2nm chips, putting European fabs near the cutting edge. Given the continent’s low market share in advanced logic, this is unlikely. More plausible is that Europe will convince a big foreign chip firm, like Intel, to build a new facility providing a stable source of supply for European automakers. Singapore continues to provide substantial incentives for chipmaking, recently winning a $4 billion investment from U.S.-based GlobalFoundries for a new fab. Japan, meanwhile, is heavily subsidizing TSMC to build a new chipmaking facility in partnership with Sony. Japan has lost much of its chipmaking in the decades since executives like Akio Morita left the scene, but Sony still retains a sizeable and profitable business making semiconductors that can sense images and which are used in the cameras in many consumer devices. Japan’s decision to subsidize a new TSMC facility, though, wasn’t primarily to help Sony. Japan’s government feared that if manufacturing kept shifting offshore, the parts of the supply chain in which Japan retains a strong position, like machine tools and advanced materials, would shift abroad, too.

虽然日本可以使用新的 Akio Morita,但美国迫切需要新的 Andy Grove。美国在芯片行业仍然拥有令人羡慕的地位。它对包括软件和机械在内的许多行业瓶颈的控制与以往一样强大。像英伟达这样的公司看起来很可能在人工智能等计算趋势的未来发挥基础性作用。此外,在芯片初创公司过时十年之后,过去几年硅谷已经将资金投入到设计新芯片的无晶圆厂公司中,这些公司通常专注于为人工智能应用优化的新架构。

While Japan could use a new Akio Morita, the United States is in desperate need of a new Andy Grove. America still has an enviable position in the chip industry. Its control over many of the industry’s choke points, including software and machinery, is as strong as ever. Companies like Nvidia look likely to play a foundational role in the future of computing trends like artificial intelligence. Moreover, after a decade in which chip startups were out of fashion, in the past few years Silicon Valley has poured money into fabless firms designing new chips, often focused on new architectures that are optimized for artificial intelligence applications.

然而,在制造这些芯片方面,美国目前落后。美国先进制造业的主要希望是英特尔。经过多年的漂泊,该公司于 2021 年任命 Pat Gelsinger 为首席执行官。Gelsinger 出生于宾夕法尼亚州的小镇,在英特尔开始了他的职业生涯,并得到了安迪·格罗夫的指导。他最终离开,在两家云计算公司担任高级职务在他被带回来扭转英特尔之前。他制定了一项雄心勃勃且成本高昂的战略,其中包含三个方面。首先是重新夺回制造领导地位,超越三星和台积电。为此,Gelsinger 与 ASML 达成协议,让英特尔获得第一台下一代 EUV 机器,预计 2025 年准备就绪。如果英特尔能够在竞争对手之前学习如何使用这些新工具,它可以提供技术边缘。

When it comes to making these chips, however, the U.S. currently lags behind. The primary hope for advanced manufacturing in the United States is Intel. After years of drift, the company named Pat Gelsinger as CEO in 2021. Born in small-town Pennsylvania, Gelsinger started his career at Intel and was mentored by Andy Grove. He eventually left to take on senior roles at two cloud computing companies before he was brought back to turn Intel around. He’s set out an ambitious and expensive strategy with three prongs. The first is to regain manufacturing leadership, overtaking Samsung and TSMC. To do this, Gelsinger has cut a deal with ASML to let Intel acquire the first next-generation EUV machine, which is expected to be ready in 2025. If Intel can learn how to use these new tools before rivals, it could provide a technological edge.

Gelsinger 战略的第二个方面是启动代工业务,与三星和台积电直接竞争,为无晶圆厂公司生产芯片,帮助英特尔赢得更多市场份额。英特尔在美国和欧洲的新设施上投入巨资,以建设潜在的未来代工客户所需的产能。然而,要让代工业务在财务上可行,可能需要赢得一些在技术前沿进行生产的客户——这意味着英特尔的代工业务只有在公司能够减少与三星和台积电的技术滞后的情况下才能发挥作用。英特尔转向代工是因为其在数据中心芯片的市场份额持续下降,这既是因为来自 AMD 和 Nvidia 的竞争,也是因为亚马逊网络服务和谷歌等云计算公司正在设计自己的芯片。

The second prong of Gelsinger’s strategy is launching a foundry business that will compete directly with Samsung and TSMC, producing chips for fabless firms and helping Intel win more market share. Intel’s spending heavily on new facilities in the U.S. and Europe to build capacity that potential future foundry customers will require. However, making the foundry business financially viable will likely require winning some customers who are producing at the technological cutting edge—meaning that Intel’s foundry business will only work if the company can reduce its technological lag with Samsung and TSMC. Intel’s foundry pivot comes as its market share in data center chips continues to decline, both because of competition from AMD and Nvidia and because cloud computing companies like Amazon Web Services and Google are designing their own chips.

英特尔的成功与否,将取决于它能否执行Gelsinger的战略,以及三星或台积电是否会出现失误。摩尔定律要求这些公司每隔几年就推出新技术,因此英特尔的一个或两个竞争对手很容易面临重大延误。然而,英特尔的战略还有一个令人不安的第三个方面:从台积电那里获得帮助。在公开场合,英特尔正在鼓励新一波芯片民族主义和对依赖亚洲生产的紧张情绪。它正试图从美国和欧洲政府那里获得补贴,以便在国内建造晶圆厂。“世界需要一个更加平衡的供应链,”Gelsinger 说。“上帝决定了石油储备在哪里,我们可以决定晶圆厂在哪里。” 然而,尽管英特尔试图整理其内部芯片制造,但它正在将越来越多的先进芯片设计的生产外包给台积电在台湾最先进的设施。

Whether Intel succeeds or fails will depend on whether it can execute Gelsinger’s strategy and whether Samsung or TSMC slip up. Moore’s Law requires these companies to roll out new technologies every few years, so one or both of Intel’s competitors could easily face major delays. Yet Intel’s strategy has an uncomfortable third prong: get help from TSMC. Publicly, Intel is encouraging a new wave of chip nationalism and nervousness about reliance on production in Asia. It’s trying to extract subsidies from both the U.S. and European governments to build fabs at home. “The world needs a more balanced supply chain,” Gelsinger argues. “God decided where the oil reserves are, we get to decide where the fabs are.” Yet while Intel tries to sort out its in-house chip fabrication, it is outsourcing production of a growing share of its advanced chip designs to TSMC’s most advanced facilities in Taiwan.

随着美国政府开始考虑先进芯片制造在东亚的集中度,美国政府说服台积电和三星在美国开设新工厂,台积电计划在亚利桑那州建立新工厂,三星在德克萨斯州奥斯汀附近扩建工厂。这些晶圆厂的部分目的是为了安抚美国政客,尽管它们还将生产用于国防和其他美国更愿意在陆上制造的关键基础设施的芯片。然而,两家公司都计划将其绝大多数生产能力以及最先进的技术留在国内。即使是美国政府的补贴承诺也不太可能改变这一点。

As it began to reckon with the concentration of advanced chipmaking in East Asia, the U.S. government convinced both TSMC and Samsung to open new facilities in the U.S., with TSMC planning a new fab in Arizona and Samsung expanding a facility near Austin, Texas. These fabs are partially intended to appease American politicians, though they will also produce chips for defense and other critical infrastructure that the U.S. would prefer to fabricate onshore. However, both companies plan to keep the vast majority of their production capacity—and their most advanced technology—at home. Even promises of subsidies from the U.S. government are unlikely to change this.

在美国国家安全官员中,关于是否以芯片设计软件和制造设备的出口管制威胁来迫使台积电在美国和台湾同时推出其最新工艺技术的讨论越来越多。或者,台积电可能被迫承诺在台湾的每一美元资本支出将与台积电在日本、亚利桑那州或新加坡的新工厂之一的资本支出相匹配。这些举措可能会开始减少世界对台湾芯片制造的依赖。但就目前而言,华盛顿不愿施加必要的压力。因此,整个世界对台湾的依赖不断增加。

Among American national security officials, there is growing discussion about whether to use threats of export controls on chip design software and manufacturing equipment to pressure TSMC to roll out its newest process technologies simultaneously in the U.S. and in Taiwan. Alternatively, TSMC could be pressed to commit that every dollar of capital expenditure in Taiwan will be matched, for example, by a dollar of capital expenditure at one of TSMC’s new facilities in Japan, Arizona, or Singapore. Such moves might begin to reduce the world’s reliance on chipmaking in Taiwan. But for now, Washington is unwilling to exert the pressure that would be required. The entire world’s dependence on Taiwan, therefore, continues to grow.

第 54 章 台湾困境

CHAPTER 54 The Taiwan Dilemma

当中国不时威胁要“对台湾开战”时,一位金融分析师问台积电董事长马克·刘:“你的客户担心吗CEO 习惯于在季度财报电话会议上提出尖锐的问题,但他们通常是关于未达到利润目标或产品发布出错的问题。在 2021 年 7 月 15 日召开本次电话会议时,台积电的财务状况看起来不错。该公司经受住了其第二大客户华为的制裁,对其业绩几乎没有任何影响。台积电股价接近历史新高。全球半导体短缺使其业务更加有利可图。曾一度在2021年成为亚洲市值最高的上市公司,全球十大市值最高的上市公司之一。

“Are your customers concerned,” one financial analyst asked TSMC chairman Mark Liu, when China from time to time threatens “a war against Taiwan?” CEOs are used to tough questions on quarterly earnings calls, but they’re usually about missed profit targets or product launches gone wrong. At the time of this call, July 15, 2021, TSMC’s financials looked fine. The company had weathered the sanctioning of its second-largest customer, Huawei, with scarcely any impact on its performance. TSMC’s share price was near a record high. The global semiconductor shortage had made its business even more lucrative. For a time in 2021, it was the most valuable publicly traded company in Asia, one of the ten most valuable publicly traded companies in the world.

然而,台积电变得越不可或缺,风险就越大——不是对台积电的财务,而是对它的设施。即使是多年来选择忽视美中对抗严重性的投资者,也开始紧张地看着台积电沿台湾海峡西海岸排列的芯片厂地图。台积电董事长坚称没有理由担心。“至于入侵中国,让我告诉你,”他宣称,“每个人都希望拥有一个和平的台海。” 出生于台北、受过伯克利大学教育和贝尔实验室培训的刘有着无可挑剔的芯片制造记录。他评估风险的能力然而,战争的力量还有待检验。他认为,鉴于世界对“台湾半导体供应链”的依赖,台海和平“对每个国家都有好处”。没有人愿意破坏它。”

Yet the more indispensable TSMC has become, the more risk has risen—not to TSMC’s financials, but to its facilities. Even investors who for years chose to ignore the severity of the U.S.-China antagonism began looking nervously at the map of TSMC’s chip fabs, arrayed along the western coast of the Taiwan Strait. TSMC’s chairman insisted that there was no reason for concern. “As to the invasion of China, let me tell you,” he declared, “everybody wants to have a peaceful Taiwan Strait.” Taipei-born, Berkeley-educated, and Bell Labs−trained, Liu has an impeccable chipmaking record. His skill in assessing the risk of war, however, has yet to be tested. Peace in the Taiwan Strait “is to every country’s benefit,” he argued, given the world’s reliance on “the semiconductor supply chain in Taiwan. No one wants to disrupt it.”

次日,7 月 16 日,数十辆解放军 05 型两栖装甲车从中国沿海冲入大海。虽然它们看起来像坦克,但这些车辆同样能够在海滩上行驶,因为它们可以像小船一样在水中疾驰。他们将在任何解放军两栖攻击中发挥重要作用。据中国官方媒体报道,在驶入大海后,数十辆这样的车辆接近驻扎在海上的登陆舰,从水面驶上舰船,准备“长途跨海”。登陆舰向他们的目标驶去。抵达后,船头的宽门打开,两栖车辆流入水中,驶向海滩,他们边走边开枪。

The next day, July 16, dozens of People’s Liberation Army Type 05 amphibious armored vehicles stormed off the Chinese coast into the ocean. Though they look like tanks, these vehicles are equally capable of driving on beaches as they are of speeding through the water like small boats. They’d be instrumental in any PLA amphibious assault. After motoring into the ocean, dozens of these vehicles approached landing ships stationed offshore, driving from the water up onto the ships, where they prepared for “a long-distance sea-crossing,” Chinese state media reported. The landing ships steamed toward their target. Upon arrival, wide doors in the ships’ bows swung open and amphibious vehicles streamed off into the water, making their way to the beach and firing their guns as they went.

这一次,只是一个练习。随后几天,解放军在台湾海峡南北入口附近进行了其他演习。“我们必须像实战一样在场景下刻苦训练,时刻做好战斗准备,坚决维护国家主权和领土完整,”中国《环球时报》援引一名营长的话说。该报尖锐地指出,演习距离普拉塔斯岛仅三百公里,这是一个与香港和台湾等距并由后者管理的小环礁。

This time, it was just an exercise. Over the next few days, the PLA launched other drills near the north and south entrances to the Taiwan Strait. “We must train hard under scenarios just like those in real battles, be combat-ready at all times and resolutely safeguard national sovereignty and territorial integrity,” China’s Global Times newspaper quoted one battalion commander as saying. The newspaper pointedly noted that the exercises took place only three hundred kilometers from Pratas Island, a tiny atoll equidistant between Hong Kong and Taiwan and administered by the latter.

对台湾的战争有多种开始方式,但一些国防规划者认为,最有可能在孤立的普拉塔斯岛引发争端。美国国防专家最近组织的一场战争游戏设想中国军队登陆该岛,并在没有开枪的情况下占领那里的台湾小型驻军。台湾和美国将面临艰难的选择,是在无关的环礁上发动战争,还是开创中国可以分割台湾大片领土的先例,例如软意大利腊肠。“温和”的回应包括在台湾部署大量美军或对中国发动网络攻击,这两者都很容易升级为全面冲突。

There are many ways a war over Taiwan could begin, but some defense planners think a ramped-up dispute over isolated Pratas Island is the most likely. One recent war game organized by American defense experts envisioned Chinese troops landing on the island and seizing the small Taiwanese garrison there without firing a shot. Taiwan and the U.S. would face the difficult choice of starting a war over an irrelevant atoll or establishing a precedent that China can slice off chunks of Taiwanese territory like pieces of soft salami. “Moderate” responses would include stationing large numbers of U.S. troops in Taiwan or launching cyberattacks on China, both of which could easily escalate into a full-blown conflict.

五角大楼关于中国军事力量的公开报告已经确定了中国可以对台湾使用武力的多种方式。最直接——但最不可能——是诺曼底登陆日式的入侵,数百艘中国舰船驶过海峡,数千名解放军步兵登陆岸上。然而,两栖入侵的历史充满了灾难,五角大楼认为这样的行动会“削弱”解放军的能力。在任何攻击之前,中国在摧毁台湾的机场和海军设施以及电力和其他关键基础设施方面几乎没有困难,但即便如此,这仍将是一场艰苦的战斗。

The Pentagon’s public reports on Chinese military power have identified multiple ways China could use force against Taiwan. The most straightforward—but most unlikely—is a D-Day style invasion, with hundreds of Chinese ships steaming across the Strait and landing thousands of PLA infantrymen on shore. The history of amphibious invasions is littered with disasters, however, and the Pentagon judges that such an operation would “strain” the PLA’s capabilities. China would have little difficulty in knocking out Taiwan’s airfields and naval facilities as well as electricity and other critical infrastructure before any assault, but even still, it would be a tough fight.

根据五角大楼的判断,其他选项对解放军来说更容易实施。台湾单靠空中和海上的部分封锁是不可能击败的。即使美日军队联手台湾打破封锁,也很难做到。中国沿岸拥有强大的武器系统。封锁不需要完全有效地扼杀岛上的贸易。结束封锁将需要台湾及其朋友——主要是美国——使数百名中国军队瘫痪位于中国领土上的系统。一场封锁破坏行动很容易演变成一场血腥的大国战争。

Other options would be easier for the PLA to implement, in the Pentagon’s judgment. A partial air and maritime blockade would be impossible for Taiwan to defeat on its own. Even if the U.S. and Japanese militaries joined Taiwan to try and break the blockade, it would be difficult to do. China has powerful weapons systems arrayed along its shores. A blockade wouldn’t need to be perfectly effective to strangle the island’s trade. Ending a blockade would require Taiwan and its friends—mainly, the U.S.—to disable hundreds of Chinese military systems sitting on Chinese territory. A blockade-busting operation could easily spiral into a bloody great power war.

即使没有封锁,光是中国的空中和导弹行动就可以使台湾的军队瘫痪,并关闭该国的经济,而无需在地面上放一双中国靴子。几天之内,如果没有美国和日本的立即援助,中国空军和导弹部队可能会解除台湾关键军事资产——机场、雷达设施、通信枢纽等——的武装,而不会严重影响台湾的生产能力。

Even without a blockade, a Chinese air and missile campaign alone could defang Taiwan’s military and shut down the country’s economy without placing a single pair of Chinese boots on the ground. In a couple days, absent immediate U.S. and Japanese aid, Chinese air and missile forces could probably disarm key Taiwanese military assets—airfields, radar facilities, communications hubs, and the like—without severely impacting the island’s productive capacity.

台积电董事长当然是正确的,没有人想“扰乱”海峡两岸纵横交错的半导体供应链。但华盛顿和北京都希望对它们进行更多控制。认为中国会出于恶意而简单地摧毁台积电晶圆厂的想法并不能使有道理,因为中国将遭受与任何人一样多的痛苦,特别是因为美国及其朋友仍然可以使用英特尔和三星的芯片工厂。中国军队入侵并直接夺取台积电的设施也不现实。他们很快就会发现,不可替代的工具的关键材料和软件更新必须从美国、日本和其他国家获得。此外,如果中国要入侵,不太可能俘获所有台积电员工。如果中国这样做,只需要少数愤怒的工程师就可以破坏整个行动。解放军已经证明它可以在两国有争议的边界上从印度夺取喜马拉雅山峰,但夺取世界上最复杂的工厂,充满爆炸性气体、危险化学品和世界上最精密的机械——那完全是另一回事。

TSMC’s chairman is certainly right that no one wants to “disrupt” the semiconductor supply chains that crisscross the Taiwan Strait. But both Washington and Beijing would like more control over them. The idea that China would simply destroy TSMC’s fabs out of spite doesn’t make sense, because China would suffer as much as anyone, especially since the U.S. and its friends would still have access to Intel’s and Samsung’s chip fabs. Nor has it ever been realistic that Chinese forces could invade and straightforwardly seize TSMC’s facilities. They’d soon discover that crucial materials and software updates for irreplaceable tools must be acquired from the U.S., Japan, and other countries. Moreover, if China were to invade, it’s unlikely to capture all TSMC employees. If China did, it would only take a handful of angry engineers to sabotage the entire operation. The PLA’s proven it can seize Himalayan peaks from India on the two countries’ disputed border, but grabbing the world’s most complex factories, full of explosive gases, dangerous chemicals, and the world’s most precise machinery—that’s a different matter entirely.

然而,很容易想象一场事故,比如空中或海上的碰撞,可能会演变成双方都不希望发生的灾难性战争。完全有理由认为,中国可能会得出这样的结论,即在没有全面入侵的情况下施加军事压力可能会决定性地破坏美国的隐含安全保证并致命地打击台湾的士气。北京知道,台湾的防务战略是打得够久,让美国和日本赶到并提供帮助。与海峡两岸的超级大国相比,这个岛屿是如此之小,以至于除了依靠朋友之外,没有任何现实的选择。想象一下,如果北京使用其海军对一小部分进出台北的船只进行海关检查。美国将如何应对?封锁是一种战争行为,但没有人愿意先开枪。如果美国什么都不做,对台湾的战斗意志的影响可能是毁灭性的。如果中国要求台积电重启华为等中国公司的芯片制造,甚至将关键人员和技术转移到大陆,台湾能说不吗?

However, it’s easy to imagine a way that an accident, like a collision in air or at sea, could spiral into a disastrous war that neither side wants. It’s also perfectly reasonable to think China might conclude that military pressure without a full-scale invasion could decisively undermine America’s implicit security guarantee and fatally demoralize Taiwan. Beijing knows that Taiwan’s defense strategy is to fight long enough for the U.S. and Japan to arrive and help. The island is so small relative to the cross-strait superpower that there’s no realistic option besides counting on friends. Imagine if Beijing were to use its navy to impose customs checks on a fraction of the ships sailing in and out of Taipei. How would the U.S. respond? A blockade is an act of war, but no one would want to shoot first. If the U.S. did nothing, the impact on Taiwan’s will to fight could be devastating. If China then demanded that TSMC restart chip fabrication for Huawei and other Chinese companies, or even to transfer critical personnel and know-how to the mainland, would Taiwan be able to say no?

这样的一系列举措对北京来说是有风险的,但并非不可想象。中国执政党没有比控制台湾更高的目标。它的领导人不断承诺这样做。政府通过了一项“反分裂国家法”,设想了其所谓的潜在用途台湾海峡的“非和平手段”。它在两栖突击车等军事系统类型上投入巨资,这是两岸入侵所需的。它定期锻炼这些能力。分析人士一致认为,海峡两岸的军事平衡已经决定性地转向中国。就像 1996 年台海危机期间那样,美国可以简单地将整个航母战斗群驶过海峡迫使北京下台的日子已经一去不复返了。现在这样的行动对美国军舰来说将充满风险。今天,中国的导弹不仅威胁到台湾周围的美国船只,而且威胁到远至关岛和日本的基地。解放军越强大,美国就越不可能冒着战争保卫台湾。如果中国要尝试对台湾进行有限军事压力,美国比以往任何时候都更有可能

Such a series of moves would be risky for Beijing, but they wouldn’t be unthinkable. China’s ruling party has no higher goal than asserting control over Taiwan. Its leaders constantly promise to do so. The government has passed an “Anti-Secession Law” envisioning the potential use of what it calls “non-peaceful means” in the Taiwan Strait. It’s invested heavily in the type of military systems, like amphibious assault vehicles, needed for a cross-strait invasion. It exercises these capabilities regularly. Analysts uniformly agree that the military balance in the Strait has shifted decisively in China’s direction. Long gone are the days, as during the 1996 Taiwan Strait crisis, that the U.S. could simply sail an entire aircraft carrier battlegroup through the Strait to force Beijing to stand down. Now such an operation would be fraught with risk for the U.S. warships. Today Chinese missiles threaten not only U.S. ships around Taiwan but also bases as far away as Guam and Japan. The stronger the PLA gets, the less likely the U.S. is to risk war to defend Taiwan. If China were to try a campaign of limited military pressure on Taiwan, it’s more likely than ever that the U.S. might look at the correlation of forces and conclude that pushing back isn’t worth the risk.

如果中国成功地向台湾施压,让其让北京平等地进入台积电的晶圆厂,甚至是优惠地进入台积电的晶圆厂,美国和日本肯定会通过对主要来自这两个国家的先进机械和材料的出口设置新的限制来做出回应和他们的欧洲盟友。但要在其他国家复制台湾的芯片制造能力需要数年时间,同时我们仍然依赖台湾。如果是这样,我们会发现自己不仅依赖中国来组装我们的 iPhone。可以想象,北京可以影响或控制唯一具有技术能力和生产能力的晶圆厂来生产我们所依赖的芯片。

If China were to succeed in pressuring Taiwan into giving Beijing equal access—or even preferential access—to TSMC’s fabs, the U.S. and Japan would surely respond by placing new limits on the export of advanced machinery and materials, which largely come from these two countries and their European allies. But it would take years to replicate Taiwan’s chipmaking capacity in other countries, and in the meantime we’d still depend on Taiwan. If so, we’d find ourselves not only reliant on China to assemble our iPhones. Beijing could conceivably gain influence or control over the only fabs with the technological capability and production capacity to churn out the chips we depend on.

这种情况对美国的经济和地缘政治地位来说将是灾难性的。如果一场战争摧毁了台积电的晶圆厂,情况会更糟。纵横交错的亚洲和台湾海峡的世界经济和供应链都建立在这种岌岌可危的和平之上。从苹果到华为再到台积电,每家在海峡两岸投资的公司都在暗中押注和平。数万亿美元被投资在从香港到新竹的台湾海峡容易被导弹射中的公司和设施上。世界芯片产业,以及所有电子产品芯片的组装,取决于台湾海峡和华南海岸的面积比世界上除硅谷以外的任何其他地区都多。

Such a scenario would be disastrous for America’s economic and geopolitical position. It would be even worse if a war knocked out TSMC’s fabs. The world economy and the supply chains that crisscross Asia and the Taiwan Strait are predicated on this precarious peace. Every company that’s invested on either side of the Taiwan Strait, from Apple to Huawei to TSMC, is implicitly betting on peace. Trillions of dollars are invested in firms and facilities within easy missile shot of the Taiwan Strait, from Hong Kong to Hsinchu. The world’s chip industry, as well as the assembly of all the electronic goods chips enable, depends more on the Taiwan Strait and the South China coast than on any other chunk of the world’s territory except Silicon Valley.

在加州的科技中心,一切照旧并不那么令人担忧。如果发生战争或地震,硅谷的大部分知识可以很容易地迁移。这在大流行期间进行了测试,当时该地区几乎所有的工人都被告知要坐在家里。大型科技公司的利润甚至上升了。如果 Facebook 华丽的总部沉入圣安地列斯断层,该公司可能几乎不会注意到。

Business as usual is not nearly as fraught in California’s tech epicenter. Much of Silicon Valley’s knowledge could be easily relocated in case of war or earthquake. This was tested during the pandemic, when almost all the region’s workers were told to sit at home. Big tech firms’ profits even went up. If Facebook’s fancy headquarters were to sink into the San Andreas Fault, the company might barely notice.

如果台积电的晶圆厂滑入1999年台湾最后一次大地震的车笼埔断层,其影响将撼动全球经济。只需要几次爆炸,无论是故意的还是意外的,都会造成类似的损害。一些粗略的计算说明了什么是危险的。台湾生产全球 11% 的内存芯片。更重要的是,它制造了全球 37% 的逻辑芯片。电脑、电话、数据中心和大多数其他电子设备都离不开它们,所以如果台湾的晶圆厂停产,第二年我们的计算能力将减少 37%。

If TSMC’s fabs were to slip into the Chelungpu Fault, whose movement caused Taiwan’s last big earthquake in 1999, the reverberations would shake the global economy. It would only take a handful of explosions, deliberate or accidental, to cause comparable damage. Some back-of-the-envelope calculations illustrate what’s at stake. Taiwan produces 11 percent of the world’s memory chips. More important, it fabricates 37 percent of the world’s logic chips. Computers, phones, data centers, and most other electronic devices simply can’t work without them, so if Taiwan’s fabs were knocked offline, we’d produce 37 percent less computing power during the following year.

对世界经济的影响将是灾难性的。COVID 后的半导体短缺提醒人们,不仅手机和电脑需要芯片。飞机和汽车、微波炉和制造设备——所有类型的产品都将面临毁灭性的延误。大约三分之一的 PC 处理器生产,包括由苹果和 AMD 设计的芯片,将被关闭,直到可以在其他地方建造新的晶圆厂。数据中心容量的增长将大幅放缓,尤其是专注于人工智能算法的服务器,这些服务器更依赖英伟达和 AMD 等公司的台湾制造芯片。其他数据基础设施将受到更大的打击。例如,新的 5G 无线电单元需要来自多家不同公司的芯片,其中许多是台湾制造的。5G网络的推出几乎完全停止。

The impact on the world economy would be catastrophic. The post-COVID semiconductor shortage was a reminder that chips aren’t only needed in phones and computers. Airplanes and autos, microwaves and manufacturing equipment—products of all types would face devastating delays. Around one-third of PC processor production, including chips designed by Apple and AMD, would be knocked offline until new fabs could be built elsewhere. Growth in data center capacity would slow dramatically, especially for servers focused on AI algorithms, which are more reliant on Taiwan-manufactured chips from companies like Nvidia and AMD. Other data infrastructure would be hit harder. New 5G radio units, for example, require chips from several different firms, many of which are made in Taiwan. There’d be an almost complete halt to the rollout of 5G networks.

停止手机网络升级是有意义的,因为购买新手机也非常困难。大多数智能手机处理器是在台湾制造的,一部典型手机的十个或更多芯片中的许多都是在台湾制造的。汽车通常需要数百个芯片才能工作,因此我们面临的延迟要比 2021 年的短缺严重得多。当然,如果爆发战争,我们需要考虑的不仅仅是芯片。中国庞大的电子组装基础设施可能会被切断。我们必须找到其他人来把我们有组件的手机和电脑组装在一起。

It would make sense to halt cell phone network upgrades because it would be extremely difficult to buy a new phone, too. Most smartphone processors are fabricated in Taiwan, as are many of the ten or more chips that go into a typical phone. Autos often need hundreds of chips to work, so we’d face delays far more severe than the shortages of 2021. Of course, if a war broke out, we’d need to think about a lot more than chips. China’s vast electronics assembly infrastructure could be cut off. We’d have to find other people to screw together whatever phones and computers we had components for.

然而,与复制台湾的芯片制造设施相比,找到新的组装工人要容易得多——尽管那会很困难。挑战不仅仅是建造新的晶圆厂。这些设施需要训练有素的人员,除非以某种方式可以从台湾撤走许多台积电员工。即便如此,新的晶圆厂也必须备有机械设备,例如 ASML 和应用材料公司的工具。在 2021 年至 2022 年芯片短缺期间,ASML 和应用材料公司都宣布他们在生产机器方面面临延迟,因为他们无法获得足够的半导体。万一发生台湾危机,他们在获取机器所需的芯片方面将面临延误。

Yet it would be far easier to find new assembly workers—as difficult as that would be—than to replicate Taiwan’s chipmaking facilities. The challenge wouldn’t simply be building new fabs. Those facilities would need trained personnel, unless somehow many TSMC staff could be exfiltrated from Taiwan. Even still, new fabs must be stocked with machinery, like tools from ASML and Applied Materials. During the 2021–2022 chip shortage, ASML and Applied Materials both announced they were facing delays in producing machinery because they couldn’t acquire enough semiconductors. In case of a Taiwan crisis, they’d face delays in acquiring the chips their machinery requires.

换句话说,在台湾发生灾难之后,总成本将以万亿计。每年损失 37% 的计算能力可能比 COVID 大流行及其经济上灾难性的封锁造成的损失更大。重建失去的芯片制造能力至少需要五年时间。这些天,当我们展望五年后,我们希望建立 5G 网络和虚拟世界,但如果台湾下线,我们可能会发现自己难以获得洗碗机。

After a disaster in Taiwan, in other words, the total costs would be measured in the trillions. Losing 37 percent of our production of computing power each year could well be more costly than the COVID pandemic and its economically disastrous lockdowns. It would take at least half a decade to rebuild the lost chipmaking capacity. These days, when we look five years out we hope to be building 5G networks and metaverses, but if Taiwan were taken offline we might find ourselves struggling to acquire dishwashers.

台湾总统蔡英文最近在《外交事务》中辩称,台湾的芯片产业是一个“‘硅盾’,让台湾能够保护自己和其他国家免受专制政权破坏全球供应链的侵略企图。” 这是看待形势的一种非常乐观的方式。台湾的芯片产业无疑迫使美国更加重视台湾的防务。然而,如果“硅盾”不能阻止中国,台湾半导体生产的集中也会使世界经济面临风险。

Taiwan’s president Tsai Ing-wen recently argued in Foreign Affairs that the island’s chip industry is a “ ‘silicon shield’ that allows Taiwan to protect itself and others from aggressive attempts by authoritarian regimes to disrupt global supply chains.” That’s a highly optimistic way of looking at the situation. The island’s chip industry certainly forces the U.S. to take Taiwan’s defense more seriously. However, the concentration of semiconductor production in Taiwan also puts the world economy at risk if the “silicon shield” doesn’t deter China.

在 2021 年的一项民意调查中,大多数台湾人认为大陆和台湾之间的战争要么不太可能(45%)要么不可能(17%)。然而,俄罗斯对乌克兰的入侵提醒人们,仅仅因为过去几十年台湾海峡基本上是和平的,一场征服战争绝非不可想象。俄乌战争还表明,任何大规模冲突在多大程度上都将部分取决于一个国家在半导体供应链中的地位,这将影响其运用军事和经济实力的能力。

In a 2021 poll, most Taiwanese reported thinking that a war between China and Taiwan was either unlikely (45 percent) or impossible (17 percent). The Russian invasion of Ukraine, however, is a reminder that just because the Taiwan Strait has been mostly peaceful for the past few decades, a war of conquest is far from unthinkable. The Russia-Ukraine War also illustrates the extent to which any large conflict will be determined in part by a country’s position in the semiconductor supply chain, which will shape its ability to wield military and economic power.

俄罗斯的芯片产业自苏联部长肖金和泽列诺格勒成立以来就落后于硅谷,冷战结束后已经衰落,因为大多数俄罗斯客户选择停止从国内芯片制造商那里购买,并将生产外包给台积电。唯一剩下的客户是俄罗斯的国防和航天工业,它们的芯片买家不足以为国内先进的芯片制造提供资金。结果,即使是俄罗斯的高优先级国防项目也难以获得所需的芯片。例如,俄罗斯相当于 GPS 卫星的由于采购半导体的问题而造成的严重延误。

Russia’s chip industry, which lagged behind Silicon Valley since the days of Soviet minister Shokin and the founding of Zelenograd, had decayed since the Cold War ended, as most Russian customers chose to stop buying from domestic chipmakers and outsourced production to TSMC. The only remaining customers were Russia’s defense and space industries, which were not big enough buyers of chips to fund advanced chipmaking at home. As a result, even high priority defense projects in Russia struggled to acquire the chips they needed. Russia’s equivalent of GPS satellites, for example, have faced wrenching delays due to problems sourcing semiconductors.

俄罗斯在制造和获取芯片方面的持续困难解释了为什么该国在乌克兰上空被击落的无人机是满是国外的微电子。这也解释了为什么俄罗斯军队继续广泛依赖非精确制导弹药。最近对俄罗斯在叙利亚的战争的分析发现,高达 95% 的投掷弹药是非制导的。俄罗斯在袭击乌克兰后几周内就面临制导巡航导弹短缺的事实,部分原因在于其半导体行业的糟糕状况。与此同时,乌克兰从西方获得了大量的制导弹药,例如标枪反坦克导弹,这些导弹依靠超过 200 个半导体,每一个都瞄准敌方坦克。

Russia’s ongoing difficulties with fabricating and acquiring chips explains why the country’s drones shot down over Ukraine are full of foreign microelectronics. It also explains why Russia’s military continues to rely extensively on non-precision-guided munitions. A recent analysis of Russia’s war in Syria found that up to 95 percent of munitions dropped were unguided. The fact that Russia faced shortages of guided cruise missiles within several weeks of attacking Ukraine is also partly due to the sorry state of its semiconductor industry. Meanwhile, Ukraine has received huge stockpiles of guided munitions from the West, such as Javelin anti-tank missiles that rely on over 200 semiconductors each as they home in on enemy tanks.

俄罗斯对外国半导体技术的依赖为美国及其盟国提供了强大的筹码。俄罗斯入侵后,美国对俄罗斯科技、国防和电信行业的某些类型芯片的销售实施了全面限制,这与欧洲、日本、韩国和台湾的合作伙伴协调。从美国英特尔到台湾台积电的主要芯片制造商现在切断克里姆林宫。俄罗斯的制造业面临着严重的破坏,俄罗斯汽车生产的很大一部分被停产。即使在国防等敏感领域,俄罗斯工厂也在采取规避措施,例如部署据美国情报部门称,用于洗碗机的芯片进入导弹系统。俄罗斯除了削减芯片消费外别无他法,因为它今天的芯片制造能力甚至比太空竞赛的鼎盛时期还要弱。

Russia’s dependence on foreign semiconductor technology has given the United States and its allies a powerful point of leverage. After Russia invaded, the U.S. rolled out sweeping restrictions on the sale of certain types of chips across Russia’s tech, defense, and telecoms sectors, which was coordinated with partners in Europe, Japan, South Korea, and Taiwan. Key chipmakers from America’s Intel to Taiwan’s TSMC have now cut off the Kremlin. Russia’s manufacturing sector has faced wrenching disruptions, with a substantial portion of Russian auto production knocked offline. Even in sensitive sectors like defense, Russian factories are taking evasive maneuvers such as deploying chips intended for dishwashers into missile systems, according to U.S. intelligence. Russia has little recourse other than to cut its consumption of chips, because its chipmaking capabilities today are even weaker than during the heyday of the space race.

然而,鉴于北京对该行业的投资以及美国依赖的大部分芯片制造能力都在解放军导弹的射程范围内,美中之间正在兴起的冷战在半导体方面将是一场不那么不平衡的比赛。如果认为乌克兰发生的事情不可能发生在东亚,那就太天真了。看看半导体在俄乌战争中的作用,中国政府分析人士公开表示,如果美中之间的紧张局势加剧,“我们必须抓住台积电。”

The emerging Cold War between the U.S. and China, however, will be a less lopsided match when it comes to semiconductors, given Beijing’s investment in the industry and given that much of the chipmaking capacity America relies on is within easy range of PLA missiles. It would be naïve to assume that what happened in Ukraine couldn’t happen in East Asia. Looking at the role of semiconductors in the Russia-Ukraine War, Chinese government analysts have publicly argued that if tensions between the U.S. and China intensify, “we must seize TSMC.”

1954 年和 1958 年,在毛泽东的军队用大炮轰炸台湾控制的岛屿之后,第一次冷战在台湾问题上也出现了对峙。今天,台湾处于更具破坏性的中国军队的射程之内——不仅有一系列中短程导弹,还有来自海峡中国一侧的龙田和惠安空军基地的飞机,从那里只需 7 分钟的飞行时间到台湾。并非巧合,在2021 年,这些空军基地升级为新的掩体、跑道延长线和导弹防御系统。一场新的台海危机将比 1950 年代的危机危险得多。核战争的风险仍然存在,尤其是考虑到中国不断增长的核武库。但这次的战场将是数字世界跳动的心脏,而不是在一个贫困的岛屿上对峙。更糟糕的是,与 1950 年代不同的是,尚不清楚中国人民解放军最终是否会退缩。这一次,北京可能会打赌它很可能会赢。

Cold War I had its own standoffs over Taiwan, in 1954 and again in 1958, after Mao Zedong’s military barraged Taiwanese-held islands with artillery. Today Taiwan is within range of far more destructive Chinese forces—not only an array of short- and medium-range missiles but also aircraft from the Longtian and Huian airbases on the Chinese side of the Strait, from which it’s only a seven-minute flight to Taiwan. Not coincidentally, in 2021, these airbases were upgraded with new bunkers, runway extensions, and missile defenses. A new Taiwan Strait crisis would be far more dangerous than the crises of the 1950s. There’d still be the risk of nuclear war, especially given China’s growing atomic arsenal. But rather than a standoff over an impoverished island, this time the battleground would be the beating heart of the digital world. What’s worse is that unlike in the 1950s, it’s not clear the People’s Liberation Army would eventually back down. This time, Beijing might wager that it could well win.

结论

Conclusion

1958 年中国人民解放军开始炮击台湾控制的金门岛后仅五天,在酷热的达拉斯夏季,杰克·基尔比向他的同事们展示了电路的所有组件——晶体管、电阻器和电容器——可能由半导体材料制成。在那之后的四天,Jay Lathrop 第一次把车停进了德州仪器的停车场。他已经为通过光刻技术制造晶体管的工艺申请了专利,但尚未获得使他能够购买新旅行车的陆军奖。几个月前,Morris Chang 辞去了他在马萨诸塞州一家电子公司的工作,搬到了德州仪器 (TI),以消除 TI 半导体制造过程中的错误的近乎神奇的能力赢得了声誉。同年,帕特·哈格蒂 (Pat Haggerty) 被任命为德州仪器 (TI) 总裁,董事会认为他为军事系统制造电子设备的愿景比生产公司成立时创造的石油勘探仪器更好。像 Weldon Word 这样的工程师,他们正在制造“智能”武器和精确传感器所需的电子设备。

It was only five days after People’s Liberation Army forces began shelling the Taiwanese-held Quemoy Island in 1958 that, amid the sweltering Dallas summer, Jack Kilby demonstrated to his colleagues that all the components of a circuit—transistors, resistors, and capacitors—could be made from semiconductor materials. Four days after that, Jay Lathrop pulled into the Texas Instruments parking lot for the first time. He’d already filed for a patent on the process of making transistors via photolithography but had yet to receive the Army prize that enabled him to buy a new station wagon. Several months earlier, Morris Chang had left his job at a Massachusetts electronics firm and moved to Texas Instruments, earning a reputation for a nearly magical ability to eliminate errors from TI’s semiconductor fabrication processes. That same year, Pat Haggerty was named president of Texas Instruments, with the board of directors betting that his vision of building electronics for military systems was a better business than producing the oil exploration instruments that the company had been founded to create. Haggerty had already assembled a talented team of engineers like Weldon Word, who were building the electronics needed for “smart” weapons and accurate sensors.

得克萨斯州与台湾在世界的另一端,但基尔比发明集成电路并非巧合中美危机。国防资金流入电子公司。美国军方依靠技术来保持其优势。随着苏联和共产主义中国建立工业规模的军队,美国不能指望部署更大的军队或更多的坦克。它可以制造更多的晶体管、更精确的传感器和更有效的通信设备,所有这些最终都会使美国的武器更加强大。

Texas was on the opposite side of the world from Taiwan, but it wasn’t a coincidence that Kilby invented his integrated circuit amid a U.S.-China crisis. Defense dollars were flowing into electronics firms. The U.S. military was relying on technology to preserve its edge. With Soviet Russia and Communist China building industrial-scale militaries, the U.S. couldn’t count on fielding bigger armies or more tanks. It could build more transistors, more precise sensors, and more effective communications equipment, all of which would eventually make American weapons far more capable.

莫里斯·张在得克萨斯州而不是在天津找工作也不是巧合。对于一个上流家庭的雄心勃勃的孩子来说,留在中国有被骚扰甚至死亡的风险。在冷战时期的混乱和席卷全球的非殖民化进程中,来自许多国家的最优秀和最聪明的人试图前往美国。John Bardeen 和 Walter Brattain 发明了第一个晶体管,但他们的贝尔实验室同事 Mohamed Atalla 和 Dawon Kahng 设计了一种可以大规模生产的晶体管结构。与 Bob Noyce 一起创立仙童半导体的“八叛徒”工程师中有两位出生在美国境外。几年后,

Nor was it a coincidence that Morris Chang was seeking work in Texas rather than, say, Tianjin. For an ambitious child of an upper-class family, staying in China risked harassment or even death. Amid Cold War chaos and the disruptions of decolonization that swept the world, the best and the brightest from many countries tried to make their way to the United States. John Bardeen and Walter Brattain invented the first transistor, but it was their Bell Labs colleagues Mohamed Atalla and Dawon Kahng who devised a transistor structure that could be mass-produced. Two of the “traitorous eight” engineers who founded Fairchild Semiconductor with Bob Noyce were born outside the United States. A few years later, a sharp-elbowed Hungarian émigré formerly known as Andras Grof helped Fairchild optimize the use of chemicals in the company’s chipmaking processes and set himself on a path to becoming CEO.

在世界上大多数人从未听说过硅芯片,更不了解它们是如何工作的时候,美国的半导体生产中心正在将世界上最聪明的人才吸引到德克萨斯州、马萨诸塞州,尤其是加利福尼亚州。这些工程师和物理学家的信念是,微型化晶体管可以完全改变未来。他们被证明远远超出了他们最疯狂的梦想。像戈登摩尔和加州理工学院教授卡弗米德这样的有远见的人看到了未来几十年,但摩尔从 1965 年开始的预测“家用电脑”和“个人便携式通信设备”几乎没有开始描述芯片在我们今天生活中的核心地位。半导体行业最终将每天生产更多晶体管的想法比有人体内的细胞对于硅谷的创始人来说是不可思议的。

At a time when most of the world had never heard of silicon chips, and still fewer understood anything about how they worked, America’s centers of semiconductor production were drawing the world’s most brilliant minds to Texas, Massachusetts, and above all to California. These engineers and physicists were driven by the belief that miniaturizing transistors could quite literally change the future. They were proven right far beyond their wildest dreams. Visionaries like Gordon Moore and Caltech professor Carver Mead saw decades ahead, but Moore’s prediction from 1965 of “home computers” and “personal portable communications equipment” barely begins to describe the centrality of chips in our lives today. The idea that the semiconductor industry would eventually produce more transistors each day than there are cells in the human body was something the founders of Silicon Valley would have found inconceivable.

随着行业规模扩大,晶体管规模缩小,对广阔的全球市场的需求比以往任何时候都更加重要。今天,即使是五角大楼的 7000 亿美元预算,也不足以负担在美国本土建造用于国防目的的尖端芯片的设施。国防部为价值 10 亿美元的潜艇和价值 100 亿美元的航空母舰专门建造了造船厂,但它使用的许多芯片都是从商业供应商那里购买的,通常是在台湾。即使是设计一个可能超过 1 亿美元的尖端芯片的成本,对于五角大楼来说也变得过于昂贵。制造最先进逻辑芯片的设施的成本是航空母舰的两倍,但其尖端技术只能维持几年。

As the industry’s scaled up, and transistors have scaled down, the need for vast, global markets is more important than ever. Today, even the Pentagon’s $700 billion budget isn’t big enough to afford facilities for building cutting-edge chips for defense purposes on U.S. soil. The Defense Department has dedicated shipyards for billion-dollar submarines and ten-billion-dollar aircraft carriers, but it buys many of the chips it uses from commercial suppliers, often in Taiwan. Even the cost of designing a leading-edge chip, which can exceed $100 million, is getting too expensive for the Pentagon. A facility to fabricate the most advanced logic chips costs twice as much as an aircraft carrier but will only be cutting-edge for a couple of years.

产生计算能力的惊人复杂性表明,硅谷不仅仅是一个科学或工程的故事。只有找到市场,技术才会进步。半导体的历史也是销售、营销、供应链管理和降低成本的故事。如果没有创建它的企业家,硅谷就不会存在。鲍勃·诺伊斯 (Bob Noyce) 是一名受过麻省理工学院训练的物理学家,但他以商人的身份取得了成功,他意识到一种尚不存在的产品的巨大市场。仙童半导体“将更多组件塞进集成电路”的能力——正如戈登·摩尔在他 1965 年的著名文章中所说——不仅取决于公司的物理学家和化学家,还取决于像查理·斯波克这样的勤奋的制造老板。追求无工会的晶圆厂并为大多数员工提供股票期权,不断提高生产力。由于现在被遗忘的 Fairchild 员工在离开公司时在离职调查中写道:“我……想要……变得……富有。”

The staggering complexity of producing computing power shows that Silicon Valley isn’t simply a story of science or engineering. Technology only advances when it finds a market. The history of the semiconductor is also a story of sales, marketing, supply chain management, and cost reduction. Silicon Valley wouldn’t exist without the entrepreneurs who built it. Bob Noyce was an MIT-trained physicist, but he made his mark as a businessman, perceiving a vast market for a product that didn’t yet exist. Fairchild Semiconductor’s ability to “cram more components onto integrated circuits”—as Gordon Moore put it in his famous 1965 article—depended not only on the company’s physicists and chemists, but also on hard-driving manufacturing bosses like Charlie Sporck. Pursuing union-free fabs and offering stock options to most employees drove productivity relentlessly higher. Transistors today cost far less than a millionth of their 1958 price thanks to the spirit expressed by the now-forgotten Fairchild employee who wrote on his exit survey when leaving the company: “I… WANT… TO… GET… RICH.”

回想起来,说芯片创造了现代世界太简单了,因为我们的社会和政治已经构建了芯片的研究、设计、生产、组装和使用方式。例如,五角大楼的研发部门 DARPA 从字面上塑造了通过资助对最先进的逻辑芯片中使用的 3D 晶体管结构(称为 FinFET)的关键研究。而未来中国的大量补贴将深刻重塑半导体供应链,无论中国是否实现了半导体霸主的目标。

On reflection, it’s too simple to say that the chip made the modern world, because our society and our politics have structured how chips were researched, designed, produced, assembled, and used. For example, DARPA, the Pentagon’s R&D unit, has literally shaped the semiconductor by funding crucial research into the 3D transistor structures, called FinFETs, used in the most advanced logic chips. And in the future China’s deluge of subsidies will profoundly reshape the semiconductor supply chain, whether China achieves its goal of semiconductor supremacy or not.

当然,不能保证芯片会像过去一样重要。我们对计算能力的需求不太可能减少,但我们可能会供不应求。戈登摩尔著名的定律只是一个预测,而不是物理事实。从 Nvidia 首席执行官 Jensen Huang 到前斯坦福大学校长和 Alphabet 董事长 John Hennessy 的行业名人都有宣布摩尔定律死亡。在某些时候,物理定律将使晶体管无法进一步缩小。甚至在此之前,制造它们可能变得过于昂贵。成本下降的速度已经明显放缓。制造更小的芯片所需的工具非常昂贵,尤其是每台成本超过 1 亿美元的 EUV 光刻机。

There’s no guarantee, of course, that chips will remain as important as they’ve been in the past. Our demand for computing power is unlikely ever to diminish, but we could run out of supply. Gordon Moore’s famous law is only a prediction, not a fact of physics. Industry luminaries from Nvidia CEO Jensen Huang to former Stanford president and Alphabet chairman John Hennessy have declared Moore’s Law dead. At some point, the laws of physics will make it impossible to shrink transistors further. Even before then, it could become too costly to manufacture them. The rate of cost declines has already significantly slowed. The tools needed to make ever-smaller chips are staggeringly expensive, none more so than the EUV lithography machines that cost more than $100 million each.

摩尔定律的终结对半导体行业乃至整个世界都是毁灭性的。我们每年生产更多的晶体管只是因为这样做在经济上是可行的。然而,这并不是摩尔定律第一次被宣布濒临死亡。1988 年,IBM 受人尊敬的专家、后来的美国国家科学基金会负责人 Erich Bloch 宣布,当晶体管缩小到四分之一微米时,摩尔定律将停止工作——这是该行业的一个障碍。十年后被猛烈抨击。Gordon Moore 在 2003 年的一次演讲中担心“一切照旧在未来十年左右肯定会遇到障碍”,但所有这些潜在的障碍都被克服了。当时,摩尔认为 3D 晶体管结构是“激进的想法”,但不到二十年后,我们已经生产了数万亿个这种 3D FinFET 晶体管。创造“摩尔定律”一词的加州理工学院教授卡弗·米德(Carver Mead)在半个世纪前就预言芯片最终可能每平方厘米包含 1 亿个晶体管,震惊了世界半导体科学家。今天,最先进的晶圆厂可以挤压一个芯片上的晶体管数量是米德认为可能的一百倍。

The end of Moore’s Law would be devastating for the semiconductor industry—and for the world. We produce more transistors each year only because it’s economically viable to do so. This isn’t the first time, though, that Moore’s Law has been declared near dead. In 1988, Erich Bloch, an esteemed expert at IBM and later head of the National Science Foundation, declared that Moore’s Law would stop working when transistors shrank to a quarter of a micron—a barrier that the industry bashed through a decade later. Gordon Moore worried in a 2003 presentation that “business as usual will certainly bump up against barriers in the next decade or so,” but all these potential barriers were overcome. At the time, Moore thought a 3D transistor structure was a “radical idea,” but less than two decades later, we’ve already produced trillions of these 3D FinFET transistors. Carver Mead, the Caltech professor who coined the phrase “Moore’s Law,” shocked the world’s semiconductor scientists with his prediction half a century ago that chips might eventually contain 100 million transistors per square centimeter. Today, the most advanced fabs can squeeze a hundred times as many transistors on a chip than even Mead thought possible.

换句话说,摩尔定律的持久性甚至令它命名的人和创造它的人都感到惊讶。今天的悲观主义者也可能对此感到惊讶。明星半导体设计师吉姆·凯勒(Jim Keller)因在苹果、特斯拉、AMD 和英特尔的芯片变革性工作而广受赞誉,他说他看到将晶体管封装在芯片上的密度提高 50 倍的明确道路。首先,他认为,现有的鳍状晶体管可以印刷得更薄,以便将三倍的数量封装在一起。接下来,鳍形晶体管将被新的管形晶体管取代,通常称为“环栅”。这些是线形管,可以从各个方向(顶部、侧面和底部)施加电场,从而更好地控制“开关”,以应对晶体管缩小带来的挑战。凯勒认为,这些细线将使晶体管的封装密度增加一倍。他预测,将这些电线堆叠在一起可以将密度进一步提高八倍。这样一来,可以容纳在芯片上的晶体管数量大约增加了 50 倍。“我们没有用完原子,”凯勒说。“我们知道如何打印单层原子。”

The durability of Moore’s Law, in other words, has surprised even the person who it’s named after and the person who coined it. It may well surprise today’s pessimists, too. Jim Keller, the star semiconductor designer who’s widely credited for transformative work on chips at Apple, Tesla, AMD, and Intel, has said he sees a clear path toward a fifty times increase in the density with which transistors can be packed on chips. First, he argues, existing fin-shaped transistors can be printed thinner to allow three times as many to be packed together. Next, fin-shaped transistors will be replaced by new tube-shaped transistors, often called “gate-all-around.” These are wire-shaped tubes that let an electric field be applied from all directions—top, sides, and bottom—providing better control of the “switch” to cope with challenges as transistors shrink. These tiny wires will double the density at which transistors can be packed, Keller argues. Stacking these wires on top of each other can increase density eight times further, he predicts. This adds up to a roughly fifty times increase in the number of transistors that can fit on a chip. “We’re not running out of atoms,” Keller has said. “We know how to print single layers of atoms.”

对于摩尔定律终结的所有讨论,流入芯片行业的资金比以往任何时候都多。过去几年,为人工智能算法设计优化芯片的初创公司筹集了数十亿美元,每个人都希望自己能成为下一个英伟达。大型科技公司——谷歌、亚马逊、微软、苹果、Facebook、阿里巴巴等——现在都在投入资金设计自己的芯片。显然没有创新不足。

For all the talk of Moore’s Law ending, there’s more money than ever before flowing into the chip industry. Startups designing chips optimized for AI algorithms have raised billions of dollars in the past few years, each hoping that they can become the next Nvidia. Big tech firms—Google, Amazon, Microsoft, Apple, Facebook, Alibaba, and others—are now pouring money into designing their own chips. There’s clearly no deficit of innovation.

支持摩尔定律即将终结的论点的最佳论据是,所有这些针对特定目的,甚至是针对个别公司的芯片新活动,正在取代英特尔经常以越来越多的节奏进行的“通用”计算的改进。 - 为过去半个世纪提供的强大微处理器。两位研究人员 Neil Thompson 和 Svenja Spanuth 甚至认为我们正在看到“计算机作为通用技术的衰落”。他们认为计算的未来将分为“获得强大的定制芯片的‘快车道’应用程序和使用进展缓慢的通用芯片卡住的‘慢车道’应用程序”。

The best argument in favor of the thesis that Moore’s Law is ending is that all this new activity in chips for specific purposes, or even for individual companies, is displacing the improvements in “general-purpose” computing that Intel’s regular cadence of ever-more-powerful microprocessors provided for the past half century. Neil Thompson and Svenja Spanuth, two researchers, have gone so far as to argue that we’re seeing a “decline of computers as a general purpose technology.” They think the future of computing will be divided between “ ‘fast lane’ applications that get powerful customized chips and ‘slow lane’ applications that get stuck using general-purpose chips whose progress fades.”

不可否认的是,作为现代计算的主力的微处理器,正在部分地被为特定目的而制造的芯片所取代。不太清楚的是这是否是一个问题。Nvidia 的 GPU 不像英特尔微处理器那样通用,因为它们是专门为图形设计的,而且越来越多地用于人工智能。然而,英伟达和其他提供针对人工智能优化的芯片的公司已经使人工智能的实施成本大大降低,因此更广泛地使用。与十年前相比,今天的人工智能已经变得更加“通用”,这在很大程度上要归功于新的、更强大的芯片。

It’s undeniable that the microprocessor, the workhorse of modern computing, is being partially displaced by chips made for specific purposes. What’s less clear is whether this is a problem. Nvidia’s GPUs are not general purpose like an Intel microprocessor, in the sense that they’re designed specifically for graphics and, increasingly, AI. However, Nvidia and other companies offering chips that are optimized for AI have made artificial intelligence far cheaper to implement, and therefore more widely accessible. AI has become a lot more “general purpose” today than was conceivable a decade ago, largely thanks to new, more powerful chips.

亚马逊和谷歌等大型科技公司最近设计自己的芯片的趋势标志着近几十年来的另一个变化。亚马逊和谷歌都进入了芯片设计业务,以提高运行其公共云的服务器的效率。任何人都可以付费访问谷歌云上的谷歌 TPU 芯片。悲观的观点是把这看作是计算的一个分支“慢车道”和“快车道”。不过,令人惊讶的是,几乎任何人都可以通过购买 Nvidia 芯片或租用 AI 优化云的访问权限来访问快车道。

The recent trend of big tech firms like Amazon and Google designing their own chips marks another change from recent decades. Both Amazon and Google entered the chip design business to improve the efficiency of the servers that run their publicly available clouds. Anyone can access Google’s TPU chips on Google’s cloud for a fee. The pessimistic view is to see this as a bifurcation of computing into a “slow lane” and a “fast lane.” What’s surprising though, is how easy it is for almost anyone to access the fast lane by buying an Nvidia chip or by renting access to an AI-optimized cloud.

此外,它比以往任何时候都更容易结合不同类型的芯片。过去,设备通常只有一个处理器芯片。现在它可能有多个处理器,一些专注于一般操作,而另一些则经过优化以管理相机等特定功能。这是可能的,因为新的封装技术使高效连接芯片变得更加容易,随着处理要求或成本考虑的变化,公司可以轻松地将某些芯片换入或换出设备。大型芯片制造商现在比以往任何时候都更加关注他们的芯片将在其中运行的系统。所以重要的问题不是我们是否最终达到了摩尔定律的极限正如 Gordon Moore 最初定义的那样——每个芯片的晶体管数量呈指数增长——但我们是否已经达到了芯片可以经济高效地生产的计算能力的峰值。成千上万的工程师和数十亿美元的赌注没有。

Moreover, it’s easier than ever before to combine different types of chips. In the past, a device would often have a single processor chip. Now it might have multiple processors, some focused on general operations, with others optimized to manage specific features like a camera. This is possible because new packaging technologies make it easier to connect chips efficiently, letting companies easily swap certain chips in or out of a device as processing requirements or cost considerations change. Big chipmakers are now putting more thought than ever before into the systems in which their chips will operate. So the important question isn’t whether we’re finally reaching the limits of Moore’s Law as Gordon Moore initially defined it—exponential increase in the number of transistors per chip—but whether we’ve reached a peak in the amount of computing power a chip can cost-effectively produce. Many thousands of engineers and many billions of dollars are betting not.

早在 1958 年 12 月,也就是 Morris Chang、Pat Haggerty、Weldon Word、Jay Lathrop 和 Jack Kilby 齐聚德州仪器的同一年,在寒冷的华盛顿特区举行了一次电子会议,当天出席的有 Chang、Gordon Moore ,和鲍勃·诺伊斯,他们都出去喝啤酒,然后,在一天即将结束的时候,年轻而兴奋地蜿蜒回到他们的旅馆,在雪堆中歌唱。他们在街上经过的任何人都不会猜到这是三位未来的科技巨头。然而,它们不仅在数十亿硅片上,而且在我们的一生中都留下了不朽的印记。他们发明的芯片和他们建立的行业提供了隐藏的电路,这些电路构成了我们的历史,并将塑造我们的未来。

Back in December 1958—the same year that saw Morris Chang, Pat Haggerty, Weldon Word, Jay Lathrop, and Jack Kilby all assembled at Texas Instruments—an electronics conference took place in a wintry Washington, D.C. Attending that day were Chang, Gordon Moore, and Bob Noyce, who all went out for beers and then, in the day’s waning hours, meandered back to their hotel, young and excited, singing amid the snowdrifts. No one they passed in the street would have guessed these were three future titans of technology. Yet they’ve left an enduring imprint not only on billions of silicon wafers but on all our lives. The chips they invented and the industry they built provide the hidden circuitry that’s structured our history and will shape our future.

图片

今天的计算机和智能手机在包含数十亿个微型晶体管的芯片上运行,这些微型电子开关可以打开和关闭以表示信息。因此,它们比 1945 年最先进的美国陆军 ENIAC 计算机功能强大得多。该设备仅包含 18,000 个“开关”。(盖蒂图片社)

Today’s computers and smartphones run on chips containing billions of microscopic transistors, the tiny electric switches that flip on and off to represent information. As such, they are unfathomably more capable than the U.S. Army’s ENIAC computer, which was state of the art for 1945. That device contained a mere 18,000 “switches.” (Getty Images)

图片

Bob Noyce()于 1957 年共同创立了飞兆半导体,其目标是制造硅晶体管。图中还有诺伊斯的长期合伙人戈登·摩尔(最左边)以及后来创立了美国最强大的风险投资公司凯鹏华盈的尤金·克莱纳(左三)。(韦恩米勒/万能照片)

Bob Noyce (center) cofounded Fairchild Semiconductor in 1957 with the goal of building silicon transistors. Also pictured is Noyce’s longtime partner Gordon Moore (far left) as well as Eugene Kleiner (third from left), who later founded Kleiner Perkins, America’s most powerful venture capital firm. (Wayne Miller/Magnum Photos)

图片

1958 年,德州仪器 (TI) 的杰克·基尔比 (Jack Kilby) 在一块半导体材料上构建了多个电子元件——第一个“集成电路”或“芯片”。(达拉斯晨报

In 1958, Jack Kilby at Texas Instruments built multiple electronic components on a single block of semiconductor material—the first “integrated circuit,” or “chip.” (Dallas Morning News)

图片

Bob Noyce 意识到推动芯片需求的是民用计算机市场,而不是军用计算机市场。他大刀阔斧地降价,以便将芯片插入民用计算机,推动该行业的发展。(泰德斯特列辛斯基/盖蒂图片社)

Bob Noyce realized it was the civilian computer market, not the military, that would drive chip demand. He aggressively cut prices so that chips could be plugged into civilian computers, fueling the industry’s growth. (Ted Streshinsky/Getty Images)

图片

德州仪器 (TI) 芯片的第一个主要订单是用于 Minuteman II 导弹上的制导计算机,如图所示。(戴夫菲尔兹)

The first major order for Texas Instruments’ chips was for the guidance computer on the Minuteman II missile, pictured here. (Dave Fields)

图片

克格勃间谍阿尔弗雷德萨兰特和乔尔巴尔都在纽约长大,他们叛逃到苏联帮助建立苏联的计算机产业。尽管苏联人偷窃,但他们未能找到尖端。(巴尔论文/史蒂文·乌斯丁)

KGB spies Alfred Sarant and Joel Barr, both of whom grew up in New York, defected to the USSR to help build the Soviet computer industry. Despite the Soviets’ pilfering, they failed to find the cutting edge. (Barr Papers/Steven Usdin)

图片

在德州仪器公司,Weldon Word 使用微电子技术制造了第一颗激光制导炸弹,该炸弹首次用于袭击越南的一座桥梁,该桥梁此前曾被数百枚“哑弹”炸弹击中。(马克珀尔斯坦/盖蒂图片社)

At Texas Instruments, Weldon Word used microelectronics to build the first laser-guided bomb, which was first used to strike a bridge in Vietnam that had previously been missed by hundreds of “dumb” bombs. (Mark Perlstein/Getty Images)

图片

1980 年代,日本向美国挑战半导体主导地位。索尼的联合创始人 Akio Morita 和 Masaru Ibuka 开创了 Sony Walkman 等变革性产品,这证明亚洲公司不仅可以高效制造,还可以赢得利润丰厚的消费市场。(索尼)

In the 1980s Japan challenged the U.S. for semiconductor dominance. Akio Morita and Masaru Ibuka, cofounders of Sony, pioneered transformative products like the Sony Walkman, which proved that Asian firms could not only manufacture effectively but also win lucrative consumer markets. (Sony)

图片

美国在亚洲的半导体组装厂为美国的盟友提供了数千个工作岗位。图为 1972 年在马来西亚槟城开设的英特尔工厂的女性。“工人主要是女性,”英特尔解释说,“因为她们在灵活性测试中表现更好。” (英特尔)

American semiconductor assembly plants across Asia provided thousands of jobs for America’s allies. Pictured here are women at an Intel facility in Penang, Malaysia, that opened in 1972. “The workers were predominantly women,” Intel explained, “because they performed better on dexterity tests.” (Intel)

图片

当张忠谋被德州仪器公司的 CEO 职位拒之门外时,他搬到了台湾,在那里他创立了台积电并建立了该国的芯片产业。台积电是亚洲最有价值的公司之一。(彭博社/盖蒂图片社)

When Morris Chang was passed over for the CEO job at Texas Instruments, he moved to Taiwan where he founded Taiwan Semiconductor Manufacturing Company and built the country’s chip industry. TSMC is one of Asia’s most valuable companies. (Bloomberg/Getty Images)

图片

面对来自亚洲的竞争,美国芯片制造商在创新上展开竞争。在戈登摩尔之后接任首席执行官的英特尔安迪格罗夫与比尔盖茨结成联盟。四十年后,微软的 Windows 软件和英特尔的 x86 芯片继续主导着 PC 业务。(美联社照片/保罗佐久间)

Facing competition from Asia, American chipmakers competed on innovation. Intel’s Andy Grove, who took over as CEO after Gordon Moore, forged an alliance with Bill Gates. Forty years later, Microsoft’s Windows software and Intel’s x86 chips continue to dominate the PC business. (AP Photo/Paul Sakuma)

图片

事实证明,这是一个非常糟糕的决定,英特尔拒绝了史蒂夫乔布斯为苹果手机制造芯片的提议。“我看不到它,”英特尔首席执行官保罗·欧德宁后来说。(卡尔·蒙东/蕉麻出版社)

In what proved to be a colossally bad decision, Intel turned down Steve Jobs’s proposal to build chips for Apple’s mobile phones. “I couldn’t see it,” Intel CEO Paul Otellini would later say. (Karl Mondon/Abaca Press)

图片

最先进的光刻机由荷兰的 ASML 制造,用于对数百万个微型晶体管进行图案化,每个晶体管都比人体细胞小得多。每台机器的成本远远超过 1 亿美元,由数十万个组件构成。(阿斯麦)

The most advanced lithography machines, which are used to pattern millions of microscopic transistors, each far smaller than a human cell, are made by ASML in the Netherlands. Each machine costs well over $100 million dollars and is built from hundreds of thousands of components. (ASML)

图片

今天,先进的芯片拥有微小的三维晶体管,每个晶体管都比冠状病毒还小,只有几纳米(十亿分之一米)宽。(IBM)

Today, advanced chips possess tiny, three-dimensional transistors, each smaller than a coronavirus, measuring a handful of nanometers (billionths of a meter) wide. (IBM)

图片

致谢

Acknowledgments

制造尖端芯片涉及数百个工艺步骤和跨越多个国家的供应链。写这本书只比制作芯片稍微简单一点。我感谢许多国家的许多人在此过程中提供了帮助。

Fabricating a cutting-edge chip involves hundreds of process steps and a supply chain that stretches across multiple countries. Writing this book was only slightly less complex than making a chip. I’m grateful to the many people, in many countries, who have helped along the way.

我感谢华盛顿特区国会图书馆的图书馆员和档案管理员提供档案材料,特别是在大流行的限制下;南方卫理公会大学;斯坦福大学; 胡佛研究所;俄罗斯科学院档案馆;和台湾中央研究院。

For making archival material available, especially amid the restrictions of the pandemic, I thank the librarians and archivists at the Library of Congress, Washington, D.C.; Southern Methodist University; Stanford University; the Hoover Institution; the Archive of the Russian Academy of Sciences; and Academica Sinica in Taiwan.

我同样感谢有机会与来自工业界、学术界和政府的半导体专家进行了一百多次采访。几十个采访对象要求在书中不透露姓名,以便他们可以自由地谈论他们的工作。然而,我要公开感谢以下个人分享见解或帮助安排采访:Bob Adams、Richard Anderson、Susie Armstrong、Jeff Arnold、David Attwood、Vivek Bakshi、Jon Bathgate、Peter Bealo、Doug Bettinger、Michael Bruck , 拉尔夫·卡尔文, 戈登·坎贝尔, 沃尔特·卡德威尔, 约翰·卡拉瑟斯, 里克·卡西迪, Anand Chandrasekher, Morris Chang, 蒋尚义, 布莱恩·克拉克, 林恩·康威, Barry Couture, Andrea Cuomo, Aart deZain Saidin, Alberto Sangiovanni-Vincentelli, Robin Saxby, Brian Shirley, Peter Simone, Marko Slusarczuk, Randy Steck, Sergey Sudjin, Will Swope, John Taylor, Bill Tobey, Roger Van Art, Dick Van Atta, Gil Varnell, Michael von Borstel, Stephen Welby、Lloyd Whitman、Pat Windham、Alan Wolff、Stefan Wurm、Tony Yen、Ross Young、Victor Zhirnov 和 Annie Zhou。当然,他们都不对我得出的任何结论负责。

I am equally grateful for the opportunity to have conducted well over one hundred interviews with semiconductor experts from industry, academia, and government. Several dozen interview subjects asked not to be named in the book so that they could speak freely about their work. I would, however, like to publicly thank the following individuals for sharing insights or helping to arrange interviews: Bob Adams, Richard Anderson, Susie Armstrong, Jeff Arnold, David Attwood, Vivek Bakshi, Jon Bathgate, Peter Bealo, Doug Bettinger, Michael Bruck, Ralph Calvin, Gordon Campbell, Walter Cardwell, John Carruthers, Rick Cassidy, Anand Chandrasekher, Morris Chang, Shang-yi Chiang, Bryan Clark, Lynn Conway, Barry Couture, Andrea Cuomo, Aart de Geus, Seth Davis, Anirudh Devgan, Steve Director, Greg Dunn, Mark Durcan, John East, Kenneth Flamm, Igor Fomenkov, Gene Frantz, Adi Fuchs, Mike Geselowitz, Lance Glasser, Jay Goldberg, Peter Gordon, John Gowdy, Doug Grouse, Chuck Gwyn, Rene Haas, Wesley Hallman, David Hanke, Bill Heye, Chris Hill, David Hodges, Sander Hofman, Tristan Holtam, Eric Hosler, Gene Irisari, Nina Kao, John Kibarian, Valery Kotkin, Michael Kramer, Lev Lapkis, Steve Leibiger, Chris Mack, Chris Malachowsky, Dave Markle, Christopher McGuire, Marshall McMurran, Carver Mead, Bruno Murari, Bob Nease, Daniel Nenni, Jim Neroda, Ron Norris, Ted Odell, Sergei Osokin, Ward Parkison, Jim Partridge, Malcolm Penn, William Perry, Pasquale Pistorio, Mary Anne Potter, Stacy Rasgon, Griff Resor, Wally Rhines, Dave Robertson, Steve Roemerman, Aldo Romano, Jeanne Roussel, Rob Rutenbar, Zain Saidin, Alberto Sangiovanni-Vincentelli, Robin Saxby, Brian Shirley, Peter Simone, Marko Slusarczuk, Randy Steck, Sergey Sudjin, Will Swope, John Taylor, Bill Tobey, Roger Van Art, Dick Van Atta, Gil Varnell, Michael von Borstel, Stephen Welby, Lloyd Whitman, Pat Windham, Alan Wolff, Stefan Wurm, Tony Yen, Ross Young, Victor Zhirnov, and Annie Zhou. None of them, of course, are responsible for any conclusions I’ve drawn.

SEMI 总裁兼首席执行官 Ajit Manocha 提供了一组非常有用的介绍。半导体行业协会的 John Neuffer、Jimmy Goodrich 和 Meghan Biery 帮助我了解了他们对行业的看法。行业资深人士 Terry Daly 对他的时间非常慷慨,我很感激他的指导。麻省理工学院林肯实验室的 Bob Loynd 和 Craig Keast 非常友好地带我参观了他们的微电子设施。我还受益于一位希望保持匿名的业内技术审阅者通过 FinFET、高 k 材料和许多其他半导体基础科学细节的指导。

Ajit Manocha, president and CEO of SEMI, provided a very helpful set of introductions. John Neuffer, Jimmy Goodrich, and Meghan Biery of the Semiconductor Industry Association helped me understand their perspective on the industry. Terry Daly, an industry veteran, was extraordinarily generous with his time, and I’m grateful for his guidance. Bob Loynd and Craig Keast at MIT’s Lincoln Labs were kind enough to give me a tour of their microelectronics facility. I also benefitted from guidance through FinFETs, high-k materials, and many other details of the science underlying semiconductors by a technical reviewer in the industry who wishes to remain anonymous.

我与 Danny Crichton 和 Jordan Schneider 的一系列迷人对话塑造了我对筹码与政治交叉点的思考。乔丹和董焱看稿并帮忙我加强了它的论点。Kevin Xu 和他不可或缺的时事通讯提供了一些关于张忠谋的重要轶事,否则我会错过的。与 Sahil Mahtani、Philip Saunders 和他们的团队的一系列对话使我对中国芯片挑战的思考更加清晰。

My thinking about the intersection of chips and politics was shaped by a fascinating series of conversations with Danny Crichton and Jordan Schneider. Jordan and Dong Yan read the manuscript and helped me sharpen its arguments. Kevin Xu and his indispensable newsletter provided some crucial anecdotes about Morris Chang that I otherwise would have missed. A set of conversations with Sahil Mahtani, Philip Saunders, and their team crystalized my thinking about China’s chip challenges.

这项研究的一部分在耶鲁大学的国际安全研究中发表。我很感谢 Paul Kennedy 和 Arne Westad 给我这个机会。有机会在海军战争学院展示早期研究,我也受益匪浅,并感谢 Rebecca Lissner 的邀请。此外,胡佛研究所的历史研讨会和美国企业研究所提供了一些论坛来解决棘手的问题,这些问题磨练了我的论点。

Portions of this research were presented at International Security Studies at Yale University. I’m grateful to Paul Kennedy and Arne Westad for the opportunity. I also benefitted greatly from the opportunity to present early stage research at the Naval War College and thank Rebecca Lissner for the invitation. In addition, the Hoover Institution’s history workshop and the American Enterprise Institute provided forums for tough questions that honed my argument.

这本书大量借鉴了关于硅谷起源和计算历史的现有研究和新闻报道。我从之前从不同角度研究过这个话题的学者和记者那里学到了很多东西,他们的工作在笔记中被引用。我特别感谢 Leslie Berlin、Geoffrey Cain、Doug Fuller、Slava Gerovitch、Paul Gillespie、Philip Hanson、James Larson、David Laws、Wen-Yee Lee、Willy Shih、Denis Fred Simon、Paul Snell、David Stumpf、David Talbot、感谢 Zachary Wasserman 和 Debby Wu 与我分享他们的研究和专业知识。George Leopold 一直是当代芯片和电子行业的有用指南。Jose Moura 在该项目的早期阶段很慷慨地介绍了他的同事。Murray Scott 是一个经常获得想法和鼓励的来源。

This book drew heavily on existing research and journalism about the origins of Silicon Valley and on the history of computing. I’ve learned much from the scholars and journalists who have previously examined different angles of this topic and whose work is cited in the notes. I am especially thankful to Leslie Berlin, Geoffrey Cain, Doug Fuller, Slava Gerovitch, Paul Gillespie, Philip Hanson, James Larson, David Laws, Wen-Yee Lee, Willy Shih, Denis Fred Simon, Paul Snell, David Stumpf, David Talbot, Zachary Wasserman, and Debby Wu for sharing their research and expertise with me. George Leopold has been a helpful guide to the contemporary chip and electronics industry. Jose Moura was generous with introductions to his colleagues at an early stage of this project. Murray Scott was a frequent source of ideas and encouragement.

我感谢 Danny Gottfried、Jacob Clemente、Gertie Robinson、Ben Cooper、Claus Soong、Wei-Ting Chen、Mindy Tu、Freddy Lin、Will Baumgartner、Soyoung Oh、Miina Matsuyama、Matyas Kisiday、Zoe Huang、Chihiro Aita 和 Sara Ashbaugh帮助收集和翻译资源。Ashley Theis 在各个方面都提供了极大的帮助。史密斯理查森基金会和斯隆基金会的支持使这项研究成为可能。

I thank Danny Gottfried, Jacob Clemente, Gertie Robinson, Ben Cooper, Claus Soong, Wei-Ting Chen, Mindy Tu, Freddy Lin, Will Baumgartner, Soyoung Oh, Miina Matsuyama, Matyas Kisiday, Zoe Huang, Chihiro Aita, and Sara Ashbaugh for help collecting and translating sources. Ashley Theis has been enormously helpful across the board. Support from the Smith Richardson Foundation and the Sloan Foundation made this research possible.

我在弗莱彻学院的同事和学生为本书中的许多想法提供了共鸣板,尤其是 DanDrezner 2019 年关于“武器化独立”的研讨会。在 FPRI,Rollie Flynn、Maia Otarashvili 和 Aaron Stein 从最初阶段就支持这项研究。当我对手稿进行最后润色时,Kori Schake、Dany Pletka 和 Hal Brands 帮助美国企业研究所成为了一个知识分子之家。我在 Greenmantle 的同事为思考技术、金融、宏观经济和政治的交叉点提供了一个令人振奋的环境。我感谢 Niall Ferguson 早期对这个项目的热情。Pierpaolo Barbieri 提供了一系列有价值的介绍;Alice Han 帮助我了解中国的科技政策;以及 Stephanie Petrella 在项目初始阶段的尖锐批评。

My colleagues and students at the Fletcher School have provided a sounding board for many of the ideas in this book, in particular Dan Drezner’s 2019 workshop on “weaponized independence.” At FPRI, Rollie Flynn, Maia Otarashvili, and Aaron Stein supported this research from its earliest stages. Kori Schake, Dany Pletka, and Hal Brands helped make the American Enterprise Institute an intellectual home as I put the finishing touches on the manuscript. My colleagues at Greenmantle have provided a stimulating environment for thinking about the intersection of technology, finance, macroeconomics, and politics. I’m grateful to Niall Ferguson for his early enthusiasm about this project; Pierpaolo Barbieri for a valuable set of introductions; Alice Han for helping me understand Chinese tech policy; and Stephanie Petrella for her incisive criticism in the project’s initial stages.

与 Rick Horgan 和整个 Scribner 团队合作非常愉快。如果没有 Toby Mundy 早期对这本书的信心,它就不会成功。乔恩·希尔曼(Jon Hillman)做了一个早期的介绍,使这个项目开始运转。

Working with Rick Horgan and the entire Scribner team has been a pleasure. Without Toby Mundy’s early confidence in this book, it wouldn’t have gotten off the ground. Jon Hillman made an early introduction that set this project rolling.

最后,也是最重要的一点,我的家人在整个项目中始终给予支持。我的父母对每一章都是严厉的批评者。露西和弗拉德是任何人都可以要求的最好的保姆。Liya、Anton 和 Evie 容忍这本书打断了早上、晚上、周末、假期和育儿假。我把这本书献给他们。

Finally, and most important, my family has been invariably supportive throughout this project. My parents have been tough critics of each chapter. Lucy and Vlad have been the best babysitters anyone could ask for. Liya, Anton, and Evie have tolerated this book interrupting mornings, evenings, weekends, vacations, and parental leaves alike. I dedicate this book to them.

关于作者

About the Author

照片:克里斯·米勒,作者

© 乔治·马歇尔摄影

© PHOTOGRAPH BY GEORGE MARSHALL

CHRIS MILLER在塔夫茨大学弗莱彻学院教授国际历史。他还担任美国企业研究所的珍妮·柯克帕特里克 (Jeane Kirkpatrick) 访问研究员、外交政策研究所的欧亚区主任以及总部位于纽约和伦敦的宏观经济和地缘政治咨询公司 Greenmantle 的董事。访问他的网站ChristopherMiller.net并在 Twitter @crmiller1上关注他。

CHRIS MILLER teaches international history at Tufts University’s Fletcher School. He also serves as Jeane Kirkpatrick visiting fellow at the American Enterprise Institute, Eurasia director at the Foreign Policy Research Institute, and as a director at Greenmantle, a New York– and London-based macroeconomic and geopolitical consultancy. Visit his website at ChristopherMiller.net and follow him on Twitter @crmiller1.

SimonandSchuster.com

SimonandSchuster.com

www.SimonandSchuster.com/Authors/Chris-Miller

www.SimonandSchuster.com/Authors/Chris-Miller

图标:脸书 图标:推特 图标:Instagram@ScribnerBooks

@ScribnerBooks

笔记

Notes

介绍

Introduction

  1. 在 USS Mustin上: “USS Mustin 穿越台湾海峡”,海军新闻稿,2020 年 8 月 19 日, https://www.navy.mil/Press-Office/Press-Releases/display-pressreleases/Article/2317449/ uss-mustin-transits-the-taiwan-strait/#images-3;Sam LaGrone,“驱逐舰 USS Mustin 在与日本军舰作战后穿越台湾海峡”, USNI 新闻,2020 年 8 月 18 日, https://news.usni.org/2020/08/18/destroyer-uss-mustin-transits-taiwan - 与日本军舰一起进行海峡跟踪行动
  2. “武力统一”: “中国称美国海军最近在台湾附近航行‘极其危险’” ,海峡时报,2020 年 8 月 20 日, https://www.straitstimes.com/asia/east-asia/china-says -最新的美国海军航行在台湾附近非常危险;刘玄尊,《解放军集中军演震慑美国台独》,《环球时报》,2020年8月23日, https://www.globaltimes.cn/page/202008/1198593.shtml
  3. 芯片扼流圈:这句话是 Murray Scott 创造的,他的Zen on Tech时事通讯塑造了我对半导体地缘政治的思考。
  4. 四分之一的芯片行业收入: Antonio Varas、Raj Varadarajan、Jimmy Goodrich 和 Falan Yinug,“加强不确定时代的全球半导体供应链”,半导体行业协会,2021 年 4 月,展览 2, https://www. Semiconductors.org/wp-content/uploads/2021/05/BCG-x-SIA-Strengthening-the-Global-Semiconductor-Value-Chain-April-2021_1.pdf;按美元价值计算,手机占半导体销售额的 26%。
  5. 它购买了大多数现成的产品: “iPhone 12 和 12 Pro 拆解”, IFixit,2020 年 10 月 20 日, https: //www.ifixit.com/Teardown/iPhone+12+and+12+Pro+Teardown/137669 .
  6. 人类历史上最昂贵的工厂: “现代世界所围绕的工厂一瞥”,《经济学人》 ,2019 年 12 月 21 日。
  7. 销量超过 1 亿部: Angelique Chatman,“分析师称 Apple iPhone 12 销量已达到 1 亿部”,CNET,2021 年 6 月 30 日;Omar Sohail,“Apple A14 Bionic 凭借 118 亿个晶体管脱颖而出” , WCCFTech,2020 年 9 月 15 日。
  8. 不是 118 亿,而是 4: Isy Haas、Jay Last、Lionel Kattner 和 Bob Norman,由 David Laws 主持,“仙童微逻辑集成电路开发和推广小组的口述历史”,计算机历史博物馆,10 月 6 日, 2007, https://archive.computerhistory.org/resources/access/text/2013/05/102658200-05-01-acc.pdf;2022 年采访大卫·劳斯。
  9. 每比特 2 美分: Gordon E. Moore,“Cramming More Components on Integrated Circuits”, Electronics 38,No. 8(1965 年 4 月 19 日), https://newsroom.intel.com/wp-content/uploads/sites/ 11/2018/05/moores-law-electronics.pdf;英特尔 1103 数据来自“Memory Lane”, Nature Electronics 1(2018 年 6 月 13 日), https://www.nature.com/articles/s41928-018-0098-9
  10. 我们每年使用的新计算能力的三分之一:根据半导体行业协会的数据,2019 年 37% 的逻辑芯片是在台湾生产的;Varas 等人,“在不确定的时代加强全球半导体供应链”。
  11. 几乎所有世界上最先进的处理器芯片: Varas 等人,“在不确定的时代加强全球半导体供应链”,第 10 页。35.
  12. 通用汽车不得不关闭工厂: Mark Fulthorpe 和 Phil Amsrud,“现在预计到 2022 年全球轻型汽车生产影响”, IHS 市场,2021 年 8 月 19 日, https ://ihsmarkit.com/research-analysis/global-light-车辆生产影响-现在预期-well-into.html
  13. 全球 44% 的存储芯片: Varas 等人,“在不确定的时代加强全球半导体供应链”。
  14. 安全许可: 2022 年采访 Morris Chang。

第 1 章从钢到硅

CHAPTER 1 From Steel to Silicon

  1. 一位勤奋好学的年轻工程师:关于 Morita 的生活细节来自 Akio Morita, Made in Japan: Akio Morita and Sony (HarperCollins, 1987)。
  2. Morris Chang 的童年: Morris CM Chang,Morris CM Chang的自传(联邦出版,2018)。感谢 Mindy Tu 提供的翻译帮助。
  3. 安迪·格鲁夫(Andy Grove)经历了同样的钢铁台风:安德鲁·格鲁夫(Andrew Grove),游泳穿越(华纳图书,2002),p。52.
  4. 仪式自杀:约翰·内森(John Nathan),索尼:私人生活(霍顿·米夫林,2001 年),p。16.
  5. 悠闲的少年生活:张,莫里斯CM张自传。
  6. 寻热导弹:森田,日本制造,p. 1.
  7. 人类“计算机”: David Alan Grier, When Computers Were Human(普林斯顿大学出版社,2005 年),ch。13; 数学表项目,从 100,000 到 200,009 的整数倒数表(哥伦比亚大学出版社,1943 年)。
  8. 距离目标一千英尺以内:罗伯特·P·帕特森,美国战略轰炸调查:总结报告(美国战争部,1945 年),第 15,美国战略轰炸调查(空军大学出版社,1987 年), https://www.airuniversity.af.edu/Portals/10/AUPress/Books/B_0020_SPANGRUD_STRATEGIC_BOMBING_SURVEYS.pdf
  9. “调试”: TR Reid, The Chip(兰登书屋,2001 年),p。11.
  10. 一万八千管: Derek Cheung 和 Eric Brach,《征服电子:建立我们电子时代的天才、远见者、自大狂和恶棍》(Roman & Littlefield,2011),第 10 页。173.

第 2 章开关

CHAPTER 2 The Switch

  1. 威廉·肖克利长期以来一直认为:乔尔·舒尔金(Joel Shurkin),《天才破碎:电子时代的创造者威廉·肖克利的兴衰》(麦克米伦,2006 年)是对肖克利的最佳描述。另见 Michael Riordan 和 Lillian Hoddeson,水晶之火:信息时代的诞生(诺顿,1997 年)。
  2. 他实际上可以 看到电子: Gino Del Guercio 和 Ira Flatow,“晶体管化!” PBS,1999, https://www.pbs.org/transistor/tv/script1.html
  3. “固态阀”: Riordan 和 Hoddeson, Crystal Fire,尤其是。第 112-114 页。
  4. 在锗中汹涌澎湃:这篇关于晶体管的叙述主要借鉴了 Riordan 和 Hoddeson、 Crystal Fire以及 Cheung 和 Brach 的《征服电子》。
  5. Shockley 设计了一个开关: Cheung 和 Brach, Conquering the Electron,第 206-207 页。
  6. 替换人类大脑: Riordan 和 Hoddeson, Crystal Fire,p。165; “科学 1948:小脑细胞”,时间,1948, http://content.time.com/time/subscriber/article/0,33009,952095,00.html

第 3 章Noyce、Kilby 和集成电路

CHAPTER 3 Noyce, Kilby, and the Integrated Circuit

  1. 在《华尔街日报》上也有: Cheung 和 Brach, Conquering the Electron,p。228.
  2. 25,000 美元:同上,p. 214.
  3. 杰克·基尔比……花费: 2021 年采访拉尔夫·卡尔文;Jay W. Lathrop,1996 年由美国新泽西州皮斯卡塔韦 IEEE 历史中心的 David Morton 进行的口述历史。
  4. 从 AT&T 获得技术许可: Jack Kilby 采访 Arthur L. Norberg,Charles Babbage 研究所,1984 年 6 月 21 日,第 11-19 页, https://conservancy.umn.edu/bitstream/handle/11299/r107410/oh074jk。 pdf?sequence=1&isAllowed=y
  5. 跟踪敌方潜艇: Caleb III Pirtle, Engineering the World:Stories from the First 75 Years of Texas Instruments(南卫理公会大学出版社,2005 年),第 10 页。29.
  6. 在同一块半导体材料上: David Brock 和 David Laws,“微电路的早期历史”, IEEE 计算历史年鉴34,第 1 期(2012 年 1 月), https://ieeexplore.ieee.org/document /6109206 ; TR Reid,芯片(兰登书屋,2001 年)。
  7. 飞兆半导体: Shurkin, Broken Genius,p. 173; “戈登·摩尔”,PBS,1999, https://www.pbs.org/transistor/album1/moore/index.html;关于仙童的其他重要书籍包括 Arnold Thackray、David C. Brock 和 Rachel Jones,摩尔定律:戈登摩尔的一生,硅谷的安静革命者(基本,2015 年)和 Leslie Berlin,微芯片背后的人:罗伯特诺伊斯和硅谷的发明(牛津大学出版社,2005 年)。
  8. Noyce 实现了 Hoerni 的“平面法”: “1959:实用单片集成电路概念专利”,计算机历史博物馆, https://www.computerhistory.org/siliconengine/practical-monolithic-integrated-circuit-concept-patented/;Christophe Lecuyer 和 David Brock,微芯片制造商(麻省理工学院出版社,2010 年);Robert N. Noyce,Semiconductor Device-and-Lead Structure,美国,2981877,1959 年 7 月 30 日提交,1961 年 4 月 25 日发布, https: //patentimages.storage.googleapis.com/e1/73/1e/7404cd5ad6325c/US2981877 .pdf;Michael Riordan,“二氧化硅解决方案”, IEEE Spectrum,2007 年 12 月 1 日, https://spectrum.ieee.org/the-silicon-dioxide-solution;柏林,微芯片背后的人,第 53-81 页。
  9. 五十倍要做:柏林,微芯片背后的人,p。112.

第 4 章升空

CHAPTER 4 Liftoff

  1. “Russ 'Moon' Circling Globe”: “据报道在 SF 上空看到卫星”,《旧金山纪事报》,1957 年 10 月 5 日,p。1.
  2. 苏联太空计划引起了信任危机:罗伯特·迪瓦恩,人造卫星挑战(牛津,1993 年)。玛格丽特·奥玛拉(Margaret O'Mara)在《知识之城:冷战科学和寻找下一个硅谷》(普林斯顿大学出版社,2015 年)中塑造了我对冷战对美国科学影响的思考Audra J. Wolfe,与苏联竞争:冷战时期美国的科学、技术和国家(约翰霍普金斯大学出版社,2013 年);和 Steve Blank,“硅谷的秘密历史”,计算机历史博物馆讲座,2008 年 11 月 20 日, https://www.youtube.com/watch?v=ZTC_RxWN_xo
  3. 消耗更多电力: Eldon C. Hall,《登月之旅:阿波罗制导计算机的历史》(美国航空学会,1996 年),第 xxi,2;Paul Cerruzi,“摩尔定律的另一面:阿波罗制导计算机、集成电路和微电子革命,1962-1975”,载于 R. Lanius 和 H. McCurdy,美国宇航局太空飞行(Palgrave Macmillan,2018 年)。
  4. “看看它们是不是真的”: Hall,《月球之旅》,p。80.
  5. 使用诺伊斯集成电路的计算机: Hall, Journey to the Moon , pp. xxi, 2, 4, 19, 80, 82;汤姆·沃尔夫,“罗伯特·诺伊斯的修修补补”,《时尚先生》 ,1983 年 12 月。
  6. 2100 万美元: Robert N. Noyce,“军用设备中的集成电路”,电气和电子工程师光谱学会,1964 年 6 月;Christophe Lecuyer,“工业用硅:1960-1967 年飞兆半导体的组件设计、大规模生产和向商业市场的转移”,历史与技术16 (1999): 183;Michael Riordan,“二氧化硅解决方案,” IEEE Spectrum,2007 年 12 月 1 日,https://spectrum.ieee.org/the-silicon-dioxide-solution
  7. 打折至 15 美元: Hall,《月球之旅》,p。83.
  8. 向军方销售电子系统: Charles Phipps,“德州仪器 IC 的早期历史:个人观点”, IEEE 计算历史年鉴34,第 1 期(2012 年 1 月):37−47。
  9. 美国军方使用的每一件电子产品: Norman J. Asher 和 Leland D. Strom,“国防部在集成电路发展中的作用”,国防分析研究所,1977 年 5 月 1 日,第 54.
  10. “像救世主一样”: 2021 年对比尔·海耶的采访;2022 年采访张忠谋。
  11. 空军开始寻找新计算机: Patrick E. Haggerty,“Strategies, Tactics, and Research”, Research Management 9,No. 3(1966 年 5 月):152−153。
  12. 聚酯薄膜胶带: Marshall William McMurran,实现准确性:计算机和导弹的遗产(Xlibris 美国,2008 年),第 281.
  13. “真的没有太多选择”: 2021 年对 Bob Nease、Marshall McMurran 和 Steve Roemerman 的采访;David K. Stumpf,民兵:定义美国核战的导弹技术史(阿肯色大学出版社,2020 年),p。214; Patrick E. Haggerty,“战略、战术和研究”,研究管理9,第 3 期(1966 年 5 月):152−153;另见 Bob Nease 和 DC Hendrickson,民兵制导和控制简史(Rockwell Autonetics Defense Electronics,1995 年);McMurran,实现准确性,ch。12. 感谢 David Stumpf 与我分享 Nease 和 Henderson 的论文。
  14. 销售的所有集成电路的 20%: Asher 和 Strom,“国防部在集成电路发展中的作用”,第 10 页。83; 霍尔,《月球之旅》,p。19; “民兵是顶级半导体用户”,《航空周刊与空间技术》 ,1965 年 7 月 26 日,第 83.

第 5 章迫击炮和批量生产

CHAPTER 5 Mortars and Mass Production

  1. Jay Lathrop 参加:与 Jay Lathrop 的通信,2021;2021 年采访沃尔特·卡德威尔;2021 年采访约翰·高迪;Jay Lathrop 和 James R. Nall,美国半导体建筑公司,2890395A,1957 年 10 月 31 日提交,1959 年 6 月 9 日发布, https://patentimages.storage.googleapis.com/e2/4d/4b/8d90caa48db31b/US2890395。 pdf ; Jay Lathrop,“钻石条例引信实验室对微电路的光刻方法”, IEEE 计算历史年鉴35,第 1 期(2013 年):48-55。
  2. 杰克·基尔比(Jack Kilby)每个星期六都在踱步:与杰·拉斯罗普(Jay Lathrop)的通信,2021;2021 年采访玛丽·安妮·波特。
  3. 玛丽安妮波特花了几个月的时间: 2021 年玛丽安妮波特访谈;Mary Anne Potter,“口述历史”,晶体管博物馆,2001 年 9 月, http://www.semiconductormuseum.com/Transistors/TexasInstruments/OralHistories/Potter/Potter_Page2.htm
  4. Morris Chang 于 1958 年加入 TI: Chang, Morris Chang 的自传;“斯坦福工程英雄讲座:Morris Chang 与John L. Hennessy 总统,”斯坦福在线,YouTube 视频,2014 年 4 月 25 日,https://www.youtube.com/watch?v=wEh3ZgbvBrE
  5. 被扔掉: Morris Chang 的口述历史,Alan Patterson 采访,计算机历史博物馆,2007 年 8 月 24 日;2022 年采访张忠谋。
  6. “如果你没有被莫里斯咬过”:比尔·海耶和吉尔·瓦内尔的访谈,2021 年。
  7. 他的生产线的产量: Morris Chang 的口述历史,Alan Patterson 采访,计算机历史博物馆,2007 年 8 月 24 日。
  8. IBM 高管: Tekla S. Perry,“Morris Chang:铸造之父”,电气和电子工程师学会 Spectrum,2011 年 4 月 19 日, https://spectrum.ieee.org/at-work/tech-careers/morris- chang-foundry-father .
  9. “除非我们能够让它发挥作用”: David Laws,“A Company of Legend:Fairchild Semiconductor 的遗产”, IEEE Annals of the History of Computing 32,No. 1(2010 年 1 月):64。
  10. “一见钟情”: Charles E. Sporck 和 Richard Molay, Spinoff:A Personal History of the Industry That Changed the World(Saranac Lake Publishing,2001 年),第 71-72 页;Christophe Lecuyer,“工业用硅”:45。

第 6 章“我……想要……变得……富有”

CHAPTER 6 “I… WANT… TO… GET… RICH”

  1. 鱼雷遥测: Asher 和 Strom,“国防部在集成电路发展中的作用”,第 15 页。74.
  2. 使用“超过 95% 的电路”: Robert Noyce,“军用设备中的集成电路”, IEEE Spectrum(1964 年 6 月):71。
  3. “不常见的职业军官”: Thomas Heinrich,“冷战军械库:硅谷的军事承包”, Enterprise & Society 3,第 2 期(2002 年 6 月):269;Lecuyer,“工业用硅”:186。
  4. a Zenith 助听器: Reid, The Chip , p. 151.
  5. “冒险就是冒险”: Dirk Hanson,《新炼金术士:硅谷和微电子革命》(Avon Books,1983 年),第 10 页。93.
  6. 洛克希德遥遥领先:美国政府武装部队技术信息局,电子设备微型化调查,PV Horton 和 TD Smith,AD269 300,弗吉尼亚州阿灵顿:美国空军弹道导弹司空中研究发展司令部,1961 年,第 23、37、39 页, https://apps.dtic.mil/sti/citations/AD0269300
  7. 后来被称为摩尔定律:摩尔,“将更多的计算机塞进集成电路”。
  8. “意味着好生意”: Asher 和 Strom,“国防部在集成电路发展中的作用”,第 10 页。73; Herbert Kleiman,《集成电路:电子行业产品创新案例研究》(乔治华盛顿大学出版社,1966 年),第 3 页。57.
  9. Fairchild 甚至以低于制造成本的价格销售产品: Lecuyer,“工业用硅”:特别是。189、194、222;Kleiman,集成电路,p。212; Ernest Braun 和 Stuart Macdonald,微型革命:半导体电子的历史和影响(剑桥大学出版社,1982 年),p。114.
  10. Fairchild 芯片为 80% 提供服务: Asher 和 Strom,“国防部在集成电路开发中的作用”,第 10 页。64; 柏林,微芯片背后的人,p。138; Lecuyer,“工业用硅”:180、188。
  11. Noyce 的降价: “Charlie Sporck 的口述历史”,计算机历史博物馆,YouTube 视频,2017 年 3 月 2 日,1:11:48, https://www.youtube.com/watch? v=duMUvoKP-pk ;Asher 和 Strom,“国防部在集成电路发展中的作用”,第 3 页。73; 柏林,微芯片背后的人,p。138.
  12. “爬行的社会主义”:柏林,《微芯片背后的人》,p。120.
  13. “我……想要……变得……变得……富有”:迈克尔·马龙,英特尔三位一体(迈克尔·柯林斯,2014 年),第 3 页。31.

第 7 章苏联硅谷

CHAPTER 7 Soviet Silicon Valley

  1. 一位不速之客来到帕洛阿尔托: Y. Nosov,“ Tranzistor-Nashe Vse。K Istorii Velikogo Otkrytiya,” Elektronika,2008 年, https ://www.electronics.ru/journal/article/363 ;AF Trutko,IREX 论文,美国国会图书馆,华盛顿特区;对于“克罗瑟斯纪念馆”,请参阅斯坦福 1960 年年鉴。
  2. 1959 年的一份 CIA 报告发现: CIA,“苏联的半导体设备生产”,CIA /RR,1959 年 11 月,59-44。
  3. 对于像 Yuri Osokin 这样雄心勃勃的年轻工程师: Lev Lapkis、Valery Kotkin、Sergei Osokin 和 Sergey Sudjin 的采访,2021;关于苏联对美国出版物的研究:NS Simonov, Nesostoyavshayasya Informatsionnaya Revolyutsiya (Universitet Dmitriya Pozharskogo, 2013), pp. 206−207; “自动化老板办公室”,《商业周刊》 ,1956 年 4 月,第 59; AA Vasenkov,“ Nekotorye Sobytiya iz Istorii Mikroelekroniki ”, Virtualnyi Kompyuternyi Muzei,2010 年, https ://computer-museum.ru/books/vasenkov/vasenkov_3-1.htm ;B. Malashevich,“ Pervie Integralnie Skhemi ”, Virtualnyi Kompyuternyi Muzei,2008 年,https://www.computer-museum.ru/histekb/integral_1.htm
  4. 每当 Osokin 放下吉他时:采访 Lev Lapkis、Valery Kotkin 和 Sergey Sudjin。
  5. “烟盒大小”: AA Shokin, Ocherki Istorii Rossiiskoi Elektroniki, v. 6 (Tehnosfera, 2014), p. 520。
  6. Joel Barr 的儿子:在苏联,Sarant 的名字是 Philip Staros,而 Barr 的名字是 Joseph Berg;他们的工作细节很大程度上来自于 Steven T. Usdin, Engineering Communism(耶鲁大学出版社,2005 年)。
  7. 巴尔和萨兰特在莫斯科郊区构想出了他们自己的版本: Usdin, Engineering Communism,p。175; 西蒙诺夫, Nesostoyavshayasya Informatsionnaya Revolyutsiya,p。212. 俄罗斯微电子专家对巴尔和萨兰特的影响规模存在一些争论。他们并不是单枪匹马地创造了苏联的计算机产业,但他们显然发挥了重要作用。
  8. 1962 年 5 月 4 日: Usdin,《工程共产主义》,第 203-209 页。
  9. “这是我们的未来”: Shokin, Ocherki Istorii Rossiiskoi Elektroniki, v. 6, pp. 522−523, 531。

第 8 章“复制”

CHAPTER 8 “Copy It”

  1. “复制它”:西蒙诺夫, Nesostoyavshayasya Informatsionnaya Revolyutsiya,p。210; 另见 AA Vasenkov,“ Nekotorye Sobytiya iz Istorii Mikroelekroniki ”, Virtualnyi Kompyuternyi Muzei,2010, https ://computer-museum.ru/books/vasenkov/vasenkov_3-1.htm ;Boris Malin 文件,IREX 论文,美国国会图书馆,华盛顿特区;Shokin, Ocherki Istorii Rossiiskoi Elektroniki v. 6, p。543.
  2. 苏联交换生……报告说学得很少: ​​B. Malashevich,“ Pervie Integralnie Shemi ”, Virtualnyi Kompyuternyi Muzei,2008, https ://www.computer-museum.ru/histekb/integral_1.htm ;西蒙诺夫, Nesostoyavshayasya Informatsionnaya Revolyutsiya,p。65; Yury R. Nosov 的口述历史,Rosemary Remackle 采访,计算机历史博物馆,2012 年 5 月 17 日,第 22-23 页。
  3. 几乎在所有类型的先进制造业中都落后: Ronald Amann 等人,《苏联工业的技术水平》(耶鲁大学出版社,1977 年)。
  4. 一些使用英寸的芯片制造机械: AA Vasenkov,“ Nekotorye Sobytiya iz Istorii Mikroelekroniki ”, Virtualnyi Kompyuternyi Muzei,2010, https ://computer-museum.ru/books/vasenkov/vasenkov_3-1.htm ;BV Malin,“ Sozdanie Pervoi Otechestvennoi Mikroshemy ”, Virtualnyi Kompyuternyi Muzei,2000 年, https: //www.computer-museum.ru/technlgy/su_chip.htm 。
  5. 无法谈论他的发明: 2021 年采访谢尔盖·奥索金。

第 9 章晶体管推销员

CHAPTER 9 The Transistor Salesman

  1. 池田表现得像个“晶体管推销员”:这个关于池田访问的记述来自松山美奈翻译的日文资料;见 Nick Kapur,日本在 Anpo 之后的十字路口(哈佛大学出版社,2018 年),p. 84; Shiota Ushio, Tokyo Wa Moetaka (讲谈社, 1988); 池田慎太郎,“池田政府对欧洲的外交和‘三支柱’理论”,《广岛国际研究杂志》 13(2007 年);Kawamura Kazuhiko,战后日本的回忆,S25(历史研究小组,2020 年)。
  2. “强大的日本是更好的风险”:美国国务院历史学家办公室,“国家安全委员会报告”,David W. Mabon 编辑,美国对外关系,1955-1957 年,日本,卷XXIII,第 1 部分(美国政府印刷局,1991 年), https://history.state.gov/historicaldocuments/frus1955-57v23p1/d28;美国国务院历史学家办公室,“No. 588 国家安全委员会执行秘书(非专业人士)的说明”,载于 David W. Mabon 和 Harriet D. Schwar 编辑,美国对外关系,1952-1954 年,中国和日本,第十四卷,第 2 部分(美国政府印刷局,1985 年), https://history.state.gov/historicaldocuments/frus1952-54v14p2/d588
  3. 美国政府支持日本重生:美国国务院历史学家办公室,《国家安全委员会报告》。
  4. 把他叫到办公室,告诉他有趣的消息:鲍勃约翰斯通,我们正在燃烧:日本企业家和电子时代的锻造(基本书籍,1999 年),p。16; Makoto Kikuchi,1994 年由美国新泽西州皮斯卡塔韦 IEEE 历史中心的 William Aspray 进行的口述历史。
  5. “我的心开始怦怦直跳”:菊地诚,“在日本研发的早期,物理学家是如何爱上硅的”,HR Huff、H. Tsuya 和 U. Gosele 合编,硅材料科学和Technology, v. 1 (The Electrochemical Society, Inc., 1998), p. 126; Makoto Kikuchi,1994 年由美国新泽西州皮斯卡塔韦 IEEE 历史中心的 William Aspray 进行的口述历史;约翰斯通,我们在燃烧,p。15.
  6. “我从未见过如此多的闪光灯”: Vicki Daitch 和 Lillian Hoddeson,真正的天才:约翰·巴丁的生命与科学:两个诺贝尔物理学奖的唯一获得者(约瑟夫·亨利出版社,2002 年),第 173-174 页.
  7. 这似乎是“奇迹”:内森,索尼,p。13; 森田,日本制造,第 70-71 页。
  8. 这个国家似乎拥有一切森田,日本制造,p。1.
  9. “不可原谅的离谱”: Hyungsub Choi,“在运输中制造知识:技术实践、组织变革和美国和日本半导体工业的崛起,1948-1960 年”,博士论文,约翰霍普金斯大学,2007 年,第 113; 约翰斯通,我们在燃烧,p。十五。
  10. “公众不知道什么是可能的”:西蒙克里斯托弗合伙人,“制造欲望:1950 年代的日本电子产品工业”,哥伦比亚大学博士论文,1997 年,第 10 页。296; Andrew Pollack,“Akio Morita,索尼联合创始人和日本商业领袖,享年 78 岁,”纽约时报,1999 年 10 月 4 日。
  11. 德州仪器曾试图推销晶体管收音机: Pirtle, Engineering the World,第 73-74 页;Robert J. Simcoe,“口袋里的革命”,美国遗产20,第 2 期(2004 年秋季)。
  12. 交出 4.5%: John E. Tilton,国际技术扩散:半导体案例(布鲁金斯学会,1971 年),第 57、141、148 页;“Leo Esaki Facts”,诺贝尔基金会, https://www.nobelprize.org/prizes/physics/1973/esaki/facts/
  13. TI “将成为计算机电子领域的索尼”: Johnstone, We Were Burning,ch。1 和第 40-41 页。
  14. 600 亿美元: Kenneth Flamm,“半导体行业的国际化”,Joseph Grunwald 和 Kenneth Flamm 编辑,全球工厂:国际贸易中的外国组装(布鲁金斯学会,1985 年),第 70; Bundo Yamada,“日本电子公司的国际化战略:对亚洲新兴工业化经济体 (NIE) 的影响”,经合组织发展中心,1990 年 10 月, https://www.oecd.org/japan/33750058.pdf
  15. 向美国政府寻求帮助: Choi, Manufacturing Knowledge in Transit , pp. 191-192。
  16. “日本是一块基石”: “营销和出口:电子业务的现状”,《电子》 ,1960 年 5 月 27 日,p。95.
  17. “一个有他们历史的民族”:亨利·基辛格,“对话备忘录,华盛顿,1973 年 4 月 10 日,上午 11:13 - 下午 12:18”,布莱德利·林恩·科尔曼,David Goldman 和 David Nickles 编辑,《美国外交关系》,1969-1976 年,E-12 卷,东亚和东南亚文献,1973-1976年(政府印刷局,2010 年),https://history。 state.gov/historicaldocuments/frus1969-76ve12/d293
  18. “我们将为您报道”: 2021 年比尔·海耶访谈;2022 年采访张忠谋;J. Fred Bucy,躲避大象:J. Fred Bucy 的自传(Dog Ear Publishing,2014 年),第 92-93 页。
  19. 提前:约翰斯通,我们在燃烧,p。364.

第 10 章“晶体管女孩”

CHAPTER 10 “Transistor Girls”

  1. 晶体管女孩保罗丹尼尔斯,晶体管女孩(雄鹿,1964 年)。
  2. Sporck 专注于效率: David C. Brock 对 Eugene J. Flath 的采访,科学史研究所,2007 年 2 月 28 日。
  3. 见鬼去吧 Charlie Sporck 的口述历史,计算机历史博物馆;Sporck 和 Molay,衍生:改变世界的行业个人历史
  4. 最大化他们的生产力: Andrew Pollack,“In the Trenchs of the Chip Wars, a Struggle for Survival”,纽约时报,1989 年 7 月 2 日;Sporck 和 Molay, Spinoff,p。63; Charlie Sporck 的口述历史,计算机历史博物馆。
  5. 他们的装配线配备了女性: Glenna Matthew,《硅谷、女性和加州梦:二十世纪的性别、阶级和机会》(斯坦福大学出版社,2002 年),第 2 章。1-3。
  6. 另一个当时只能手工完成的步骤: Sporck 和 Molay, Spinoff,第 87-88 页。
  7. 很快上飞机: Sporck 和 Molay, Spinoff,第 91-93 页;William F. Finan,在质量上与日本相匹敌:美国领先的半导体公司如何赶上日本最好的公司(麻省理工学院日本计划,1993 年),p。61; Julius Blank 采访 David C. Brock,科学史研究所,2006 年 3 月 20 日,p。10个;Julius Blank 的口述历史,Craig Addison 采访,计算机历史博物馆,2008 年 1 月 25 日。
  8. “愿意容忍单调的工作”:约翰·亨德森,高科技生产的全球化(Routledge,1989 年),p。110; Sporck 和 Molay, Spinoff,p。94; Harry Sello 口述历史采访 Craig Addison,SEMI,2004 年 4 月 2 日。
  9. 加利福尼亚州的昂贵得令人望而却步: Sporck 和 Molay, Spinoff,p。95; Charlie Sporck 的口述历史,计算机历史博物馆。
  10. 只有一角钱: William F. Finan,“通过美国公司进行的半导体技术国际转让”,NBER 工作论文编号。118,1975 年 12 月,第 61-62 页。
  11. “非常违法”: Craig Addison,2004 年 6 月 17 日对 Clements E. Pausa 的口述历史访谈。
  12. “我们在东方从来没有遇到过任何工会问题”: Charlie Sporck 的口述历史,计算机历史博物馆;另见关于工会、工资谈判和国际劳工组织法规的广泛讨论计算机历史博物馆,“Fairchild 口述历史小组:制造和支持服务”,2007 年 10 月 5 日。

第11章精确打击

CHAPTER 11 Precision Strike

  1. 飞行中途: 2021 年采访比尔·海耶。
  2. 80万吨: Samuel J. Cox,“H-017-2:滚雷——简要概述”,海军历史和遗产司令部,2018 年 3 月 27 日, https ://www.history.navy.mil/about- us/leadership/director/directors-corner/h-grams/h-gram-017/h-017-2.html#:~:text=这些%20U.S.%20strikes%20dropped%20864%2C000,years% 20of%20World%20War%20II
  3. 只有四个例子:巴里·瓦茨,制导弹药和战斗网络的六个十年:进展和前景(战略和预算评估中心,2007 年),第 10 页。133.
  4. 其余的只是错过了:美国政府海军航空系统司令部,“1968 年 7 月至 11 月空对空导弹系统能力审查报告”,AD-A955-143,海军历史和遗产司令部,2021 年 4 月 23 日, https: //www.history.navy.mil/research/histories/naval-aviation-history/ault-report.html;瓦茨,制导弹药的六个十年,p。140.
  5. 距离目标 420 英尺以内: James E. Hickey,精确制导弹药和战争中的人类苦难(Routledge,2016 年),p。98.
  6. TI 已经生产了必要的组件: 2021 年采访 Steve Roemerman;Paul G. Gillespie,“精确制导弹药:制造比原子弹更强大的炸弹”,博士论文,利哈伊大学,2002 年。
  7. “便宜又熟悉”: 2021 年采访史蒂夫·罗默曼。
  8. 价格像廉价的家用轿车: 2021 年采访史蒂夫·罗默曼。
  9. 德州仪器能否提供帮助?: “小约瑟夫·戴维斯上校讣告”,西北佛罗里达每日新闻,2014 年 8 月 24-26 日;Gillespie,“精确制导弹药”,第 117-118 页;Walter J. Boyne,“Breaking the Dragon's Jaw”,空军杂志,2011 年 8 月,第 58-60 页, https: //www.airforcemag.com/PDF/MagazineArchive/Documents/2011/August%202011⁄0811jaw.pdf ; Vernon Loeb,“光彩的爆发”,华盛顿邮报,2002 年 12 月 15 日。
  10. Word 以标准发行的炸弹开头: Gillespie,“Precision Guided Munitions”,第 15 页。116.
  11. 精确破坏的工具:同上,第 125、172 页。
  12. “自动火力控制”:威廉·比彻,“越战后威斯特摩兰预见自动战”,纽约时报,1969 年 10 月 14 日。然而,防御理论家已经意识到精确弹药将改变战争;参见 James F. Digby,精确制导弹药:能力和后果,兰德论文 P-5257,1974 年 6 月,以及精确制导技术:改变武器优先级、新风险、新机遇,兰德论文 P-5537,1975 年 11 月。

第 12 章供应链治国之道

CHAPTER 12 Supply Chain Statecraft

  1. “酒吧和舞女”: “台湾的半导体发展并非一帆风顺”,tr。克劳斯·宋,风暴媒体,2019 年 6 月 5 日, https: //www.storm.mg/article/1358975?mode=whole.000 。
  2. 达拉斯警察的儿子: “Mark Shepherd Jr. Obituary”,达拉斯晨报,2009 年 2 月 6-8 日;Ashlee Vance,“Mark Shepherd,电子领域的一支力量,享年 86 岁”,《纽约时报》,2009 年 2 月 9 日。
  3. “欺负欠发达国家”: “台湾半导体发展并非一帆风顺”;2022 年采访张忠谋。
  4. 美国削减经济援助: David W. Chang,“美国对台湾的援助与经济进步”,亚洲调查5,第 3 期(1965 年 3 月):156;尼克·库拉瑟(Nick Cullather),“‘好龙的燃料’:美国和台湾的产业政策,1950−1960 年,”外交史20,第 1 期(1996 年冬季):1。
  5. 像KT Li这样的官员: Wolfgang Saxon,“台湾91岁的李国廷去世;引领经济转型,”纽约时报,2001 年 6 月 2 日。
  6. Chang的两个博士同学: “台湾的半导体发展并非一帆风顺。”
  7. 运送了第 10 亿个单位: L. Sophia Wang、 KT LI 和台湾经验(国立清华大学出版社,2006 年),第 10 页。216; “TI Taiwan Chronology”,载于Far East Briefing Book,德州仪器论文,南卫理公会大学图书馆,1989 年 10 月 18 日。
  8. “解决失业问题”:亨利·基辛格,“对话备忘录,华盛顿,1973 年 4 月 10 日,上午 11:13 - 下午 12:18”,载于 Bradley Lynn Coleman、David Goldman 和 David Nickles 合编,外交关系美国,1969-1976,E-12 卷,东亚和东南亚文献,1973-1976(政府印刷局,2010), https://history.state.gov/historicaldocuments/frus1969-76ve12/d293;Linda Lim 和 Pang Eng Fong,新加坡的贸易、就业和工业化(国际劳工局,1986 年),p。156.
  9. 雇佣了成千上万的工人: Joseph Grunwald 和 Kenneth Flamm,《全球工厂:国际贸易中的外国议会》(布鲁金斯学会出版社,1994 年),第 100。
  10. 高薪电子组装工作: Kenneth Flamm,“半导体行业的国际化”,Grunwald 和 Flamm, The Global Factory,第 1 页。110; Lim 和 Pang Eng Fong,新加坡的贸易、就业和工业化,p。156; 香港统计年鉴(政府统计处,1984),表3.12, https://www.censtatd.gov.hk/en/data/stat_report/product/B1010003/att/B10100031984AN84E0100.pdf;GT Harris 和 Tai Shzee Yew,“1970 年代马来西亚半岛的失业趋势”,东盟经济公报2,第 2 期(1985 年 11 月):118-132。
  11. “TI 将在台湾留下并继续发展”: 与李总理会面,台北,1977 年 9 月 23 日,以及接待/巴菲特-台北。1977 年 9 月 23 日。Mark Shepherd 评论,在 Mark Shepherd 论文、通信、报告、演讲中,1977 年,南卫理公会大学图书馆,文件夹 90-69;联系出版社,“小马克·谢泼德;德州仪器,”波士顿环球报,2009 年 2 月 9 日。

第 13 章英特尔的革命者

CHAPTER 13 Intel’s Revolutionaries

  1. “创始人离开仙童”: Marge Scandling,“2 位创始人离开仙童;组建自己的电子公司”, Palo Alto Times,1968 年 8 月 2 日。
  2. 磁芯跟不上: Lucien V. Auletta、Herbert J. Hallstead 和 Denis J. Sullivan,“铁氧体磁芯平面和阵列:IBM 的制造发展”, IEEE Transactions on Magnetics 5,第 4 期(1969 年 12 月);John Markoff,“IBM 的 Robert H. Dennard 和改变世界的芯片”,IBM,2019 年 11 月 7 日, https://www.ibm.com/blogs/think/2019/11/ibms-robert-h-dennard -and-the-chip-that-c​​hanged-the-world/
  3. Hoff 的计算机架构背景: Emma Neiman,“A Look at Stanford Computer Science, Part I: Past and Present”,斯坦福日报,2015 年 4 月 15 日;“Marcian E. Hoff, Jr. 访谈,1995 年 3 月 3 日,”斯坦福图书馆,1995 年 3 月 3 日, https://exhibits.stanford.edu/silicongenesis/catalog/jj158jn5943
  4. 没有人比 Intel 的内存芯片更强大: Robert N. Noyce 和 Marcian E. Hoff,“A History of Microprocessor Development at Intel”, IEEE Micro 1, No. 1(1981 年 2 月);David Laws 对 Ted Hoff 和 Stan Mazor 的采访,计算机历史博物馆,2006 年 9 月 20 日;“Ted Hoff:微处理器的诞生及超越”,斯坦福工程学院,2006 年 11 月。
  5. 掀起了一场计算革命: Sarah Fallon,“第一台微处理器的秘密历史”,《连线》 ,2020 年 12 月 23 日;Ken Shirriff,“第一个微处理器的惊人故事”, IEEE Spectrum,2016 年 8 月 30 日。
  6. “这将改变世界”:柏林,《微芯片背后的人》,p。205; Gordon Moore,“关于微处理器”, IEEE,1976;Ross Knox Bassett,迈向数字时代(约翰霍普金斯大学出版社,2002 年),p。281; Malone,《英特尔三位一体》,第 177-178 页;Gene Bylinsky,“英特尔如何在内存芯片上赢得赌注”,《财富》,1973 年 11 月;法伦,“第一台微处理器的秘史”。
  7. 掏出一只袜子: 2021 年对 Carver Mead 的采访。
  8. “从我们的耳朵里出来”: Carver Mead,“将电源放在它所属的地方的计算机”,工程与科学XXXVI,第 4 期(1972 年 2 月)。
  9. “我们确实是当今世界的革命者”: Gene Bylinsky,“英特尔如何在内存芯片上赢得赌注。”

第十四章五角大楼的抵消战略

CHAPTER 14 The Pentagon’s Offset Strategy

  1. 美国最顶尖的军事事务专家之一:威廉·佩里接受弗吉尼亚大学米勒中心的罗素·莱利的采访,2006 年 2 月 21 日;William J. Perry,我在核边缘的旅程(斯坦福安全研究,2015 年),ch。1-2。
  2. 从他的歌唱伙伴那里买筹码:采访威廉佩里,2021;Zachary Wasserman,“发明创业资本主义”,耶鲁大学博士论文,2015 年。
  3. 机床工厂: Andrew Krepinevich 和 Barry Watts,《最后的战士:安德鲁·马歇尔与现代美国国防战略的塑造》(基础书籍,2015 年),第 4、9、95 页。
  4. 计算机领域的“实质性和持久领先”: AW Marshall,“与苏联的长期竞争:战略分析框架”,兰德公司,R-862-PR,1972 年 4 月, https://www.rand.org/发布/报告/R862.html
  5. 30 至 500 亿美元: William Perry 的证词,国防部参议院军事委员会,79 财年拨款授权,第 8 部分:研究与开发,第 96 届美国国会,1979 年,第 5506-5937 页;Kenneth P. Werrell,巡航导弹的演变(航空大学出版社,1985 年),p。180。
  6. 区分鲸鱼和潜艇: Richard H. Van Atta、Sidney Reed 和 Seymour J. Deitchman,DARPA技术成就第二卷(国防分析研究所,1991 年),p. “12-2。”
  7. 像战斧这样的新系统: Werrell,巡航导弹的演变,p。136.
  8. “突击破坏者”: Van Atta 等人,DARPA技术成就第二卷,第 5-10 页。
  9. “各级‘智能’武器”:史蒂夫·罗默曼访谈,2021 年;威廉·J·佩里接受国防部长办公室阿尔弗雷德·戈德堡的采访,1981 年 1 月 9 日。
  10. “花里胡哨”:弗雷德卡普兰,“巡航导弹:神奇武器还是哑弹?” 高科技,1983 年 2 月;James Fallows,国防(兰登书屋,1981 年),p。55; William Perry,“Fallows 的谬误:一篇评论文章”,国际安全6,第 4 期(1982 年春季):179。
  11. “Luddites”: 2006 年 2 月 21 日,弗吉尼亚大学米勒中心的 Russell Riley 对 William Perry 的采访。

第15章“竞争很激烈”

CHAPTER 15 “That Competition Is Tough”

  1. “我的生活就像地狱一样”: 2021 年采访理查德·安德森;Michael Malone、 Bill 和 Dave:Hewlett 和 Packard 如何打造世界上最伟大的公司(精装投资组合,2006 年);“半导体市场条件和国际贸易”,众议院筹款委员会贸易小组委员会现场听证会,第 96 届国会众议院,1980 年 4 月 28 日。
  2. “点击,点击”:迈克尔·马龙, The Big Score(条纹出版社,2021 年),p。248; Jorge Contreras、Laura Handley 和 Terrence Yang,“Breaking New Ground in the Law of Copyright”, Harvard Law Journal of Technology 3(1990 年春季)。
  3. 十倍的糟糕:罗森电子通讯,1980 年 3 月 31 日。
  4. “长尾”: Malone,《英特尔三位一体》 ,p。284; Fred Warshofsky,芯片战争:明日世界之战(斯克里布纳,1989 年),第 3 页。101.
  5. 公司的五款尖端集成电路: TPS-L2:用户手册(Sony Corporation,1981),p. 24.
  6. 3.85亿: “卷。20:Walkman 进入全球词汇,”索尼, https://www.sony.com/en/SonyInfo/CorporateInfo/History/capsule/20/
  7. “竞争很激烈”: Charlie Sporck 的口述历史,计算机历史博物馆。

第16章“与日本交战”

CHAPTER 16 “At War with Japan”

  1. “我无法摆脱战斗”:马克·西蒙,“杰里·桑德斯/硅谷的硬汉”,旧金山纪事报,2001 年 10 月 4 日;Thomas Skornia,实现美国梦的案例研究:Sanders 和 Advanced Micro Devices:前 15 年,1969-1984 (1984), https://archive.computerhistory.org/resources/access/text/2019/01/ 102721657-05-01-acc.pdf
  2. “Knock 'em down, Fight 'em, kill 'em”: Charlie Sporck 的口述历史,计算机历史博物馆。
  3. “一场经济战争”: Michael S. Malone,“加利福尼亚州东京”,《纽约时报》,1981 年 11 月 1 日;Charlie Sporck 的口述历史,计算机历史博物馆。
  4. 日立的员工被捕: Thomas C. Hayes,“美国邮政保释作为 FBI 展开的行动细节”,纽约时报,1982 年 6 月 25 日。
  5. 更安静的潜艇: Wende A. Wrubel,“东芝-康士伯事件:Cocom 的缺点,以及提高对东集团出口管制有效性的建议”,《美国大学国际法评论》第 4 期,第 1 期(2011 年)。
  6. 肮脏交易: Stuart Auerbach,“CIA 称东芝向苏联集团出售了更多产品”,华盛顿邮报,1988 年 3 月 15 日。
  7. 日本市场份额低: Michael E. Porter 和 Mariko Sakakibara,“日本的竞争”,《经济展望》杂志18,第 1 期(2004 年冬季):36;政府瞄准对世界半导体竞争的影响(半导体工业协会,1983 年),第 69-74 页。
  8. 一半预算: Kiyonari Sakakibara,“从模仿到创新:日本的超大规模集成 (VLSI) 半导体项目”,工作论文,麻省理工斯隆管理学院,1983 年 10 月, https://dspace.mit.edu/handle /1721.1/47985
  9. “18% 的好日子”: Reid, The Chip,p。224.
  10. 导致他们破产: 政府瞄准对世界半导体竞争的影响,第 10 页。67.
  11. 支付较低的借款利率: Jeffrey A. Frankel,“1980 年代的日本金融:调查”,国家经济研究局,1991 年;来自 data.worldbank.org 的家庭储蓄、家庭消费和银行贷款占 GDP 百分比的数据。
  12. 1.7% 的全球 DRAM 市场: PR Morris, A History of the World Semiconductor Industry (Institute of Electrical Engineers, 1990), p. 104; Robert Burgelman 和 Andrew S. Grove,战略就是命运:战略制定如何塑造公司的未来(自由出版社,2002 年),第 3 页。35.
  13. 乐于买单: Scott Callan,“日本,解体:日本高科技联盟的竞争与冲突,成功与失败”,博士论文,斯坦福大学,1993 年,第 188,表 7.14;Clair Brown 和 Greg Linden,《芯片与变革:危机如何重塑半导体行业》(麻省理工学院出版社,2009 年)。

第 17 章“运输垃圾”

CHAPTER 17 “Shipping Junk”

  1. “最热门的高科技公司”: Clayton Jones,“Computerized Laser Swiftly Carves Circuits for Microchips”, Christian Science Monitor,1981 年 3 月 10 日;David E. Sanger,“对小型 GCA 的大担忧”,《纽约时报》,1987 年 1 月 19 日。
  2. Bob Noyce 上下驾驶:柏林,微芯片背后的人,第 94、119 页。感谢 Chris Mack 向我指出这一点。
  3. Perkin Elmer 的扫描仪:采访 Chris Mack,2021;2021 年采访戴夫·马克尔;Perkin Elmer,“Micralign 投影掩模对准系统”,芯片历史中心, https ://www.chiphistory.org/154-perkin-elmer-micralign-projection-mask-alignment-system ;Daniel P. Burbank,“制造微芯片几乎不可能”,《发明与技术》 (1999 年秋季);Alexis C. Madrigal,“绝密:关于 CIA 冷战间谍卫星‘大鸟’的简报,”大西洋,2011 年 12 月 29 日;Chris Mack,“光学光刻工具供应商的里程碑”, http://www.lithoguru.com/scientist/litho_history/milestones_tools.pdf
  4. 苏联的照片: Craig Addison 对 James E. Gallagher 的采访,SEMI,2005 年 3 月 9 日;Arthur W. Zafiropoulo 采访 Craig Addison,SEMI,2006 年 5 月 25 日;美国地球物理公司,“关于我们的公司成员”,美国气象学会公告,1962 年 12 月 12 日;琼斯,“计算机激光快速雕刻微芯片电路。”
  5. Morris Chang 走上前去: Griff Resor 访谈,2021;“Griff Resor on Photolithography”,Semi-History,YouTube 视频,2009 年 1 月 30 日,2:30, https://www.youtube.com/watch?v=OKfdHZCEfmY
  6. GCA 介绍了它的第一步: “Griff Resor on Photolithography”,SemiHistory,YouTube 视频,2009 年 1 月 30 日,2:30, https://www.youtube.com/watch?v= OKfdHZCEfmY ;Chris Mack,“光学光刻工具供应商的里程碑”, http://www.lithoguru.com/scientist/litho_history/milestones_tools.pdf;“截至 1980 年 9 月,GCA 伯灵顿分部的所有 4800 DSW 的发货历史”,第 3 页。1、由作者所有。
  7. 股价飙升: Rebecca Marta Henderson 的销售数据,“老牌企业在技术变革面前的失败”,哈佛大学博士论文,1988 年,第 217; 琼斯,“计算机激光快速雕刻微芯片电路。”
  8. “醉酒水手”: 2021 年对彼得·比洛、罗斯·杨和比尔·托比的采访;James E. Gallagher 采访 Craig Addison,SEMI,2005 年 3 月 9 日。
  9. “我们有米尔特”:对比尔·托比、吉姆·尼罗达和彼得·比洛的采访,2021 年;Ross Young, Silicon Sumo(半导体服务,1994 年),p。279; Charles N. Pieczulewski,“领先制造商中半导体光刻设备开发和采购实践的基准测试”,硕士论文,麻省理工学院,1995 年,p。54.
  10. “客户受够了”:对 Griff Resor、Bill Tobey、Jim Neroda 和 Peter Bealo 的采访,2021 年;杨,硅相扑,p。279.
  11. 雷暴滚滚而来:2021 年 Griff Resor 访谈。
  12. “纸上企业家主义”:罗伯特·赖希,《下一个美国前沿》(Crown,1983 年),第 10 页。159.
  13. “傲慢”和“不回应”: 2021 年 Gil Varnell 访谈;丽贝卡·玛尔塔·亨德森,“老牌企业在技术变革面前的失败”,p。225; 美国商务部出口管理局战略产业和经济安全办公室战略分析司美国半导体晶圆加工工业设备的国家安全评估(1991 年),第 4-10 页。
  14. 十倍于该持续时间: Henderson,“老牌企业在技术变革面前的失败”,第 220-222、227 页;2021 年采访 AMD 前高管。
  15. 没有扭转局面的计划: Pete Bealo 和 Bill Tobey 访谈,2021 年;Henderson,“老牌企业在技术变革面前的失败”,第 222-225 页;Jay Stowsky,“最薄弱的环节:半导体生产设备、联系和国际贸易的限制”,工作论文,加州大学伯克利分校,1987 年 9 月,第 2.
  16. 每个人都可以轻松一点: Craig Addison 对 Arthur W. Zafiropoulo 的采访,SEMI,2006 年 5 月 25 日;2021 年对 Peter Bealo 和 Jim Neroda 的采访。

第 18 章1980 年代的原油

CHAPTER 18 The Crude Oil of the 1980s

  1. 在倾斜的宝塔式屋顶下: Skornia、 Sanders 和 Advanced Micro Devices,p。138; Daryl Savage,“帕洛阿尔托:明的餐厅将于 12 月 28 日关闭”,帕洛阿尔托在线,2014 年 12 月 18 日, https://www.paloaltoonline.com/news/2014/12/18/mings-restaurant-to-close -dec-28
  2. “1980 年代的原油”: Arthur L. Robinson,“美国微电路制造商的危险时期”, Science 208,第 4444 期(1980 年 5 月 9 日):582;Skornia、 Sanders 和 Advanced Micro Devices,p。140.
  3. “半导体沙特阿拉伯”: Marvin J. Wolf,《日本的阴谋:主导全球工业的阴谋》(新英语图书馆,1984 年),p。83.
  4. “只是我们不能失去的东西”: David E. Sanger,“Big Worries Over Small GCA”,纽约时报,1987 年 1 月 19 日。
  5. “硅谷牛仔”: Richard Van Atta 访谈,2021 年。
  6. 五角大楼的工作组:国防科学委员会,国防半导体依赖性报告——1987 年 2 月,第 1-2 页。
  7. “you're in nowheresville”: Charlie Sporck 的口述历史,计算机历史博物馆。

第19章死亡螺旋

CHAPTER 19 Death Spiral

  1. “我们正处于死亡螺旋中”:柏林,微芯片背后的人,第 10 页。264.
  2. 超过 90%: Richard Langlois 和 Edward Steinmueller,“战略与环境”,工作论文,康涅狄格大学,1999 年,第 1166。
  3. 薯片: Clyde V. Prestowitz, Jr.,“超越自由放任”,外交政策,第 87 期(1992 年夏季):71;与迈克尔·博斯金(Michael Boskin)的电子邮件交流,2021;尽管这句话在许多文章中重复出现,但我没有发现他确实说过这句话的证据。
  4. 向国会作证:柏林,《微芯片背后的人》,p。262; John G. Rauch,“我们时代的现实”,福特汉姆知识产权、媒体和娱乐法杂志,第 3 期,第 2 期(1993 年):412。
  5. DRAM 对日本的销售几乎没有变化: Wolf,日本的阴谋,第 5、91 页;2021 年采访艾伦·沃尔夫;柏林,微芯片背后的人,p。270。
  6. 更高的价格实际上使日本的生产商受益: Doug Irwin,“Trade Politics and the Semiconductor Industry”,NBER 工作文件 W4745,1994 年 5 月。
  7. 创建了一个财团: Young, Silicon Sumo,第 262-263 页。
  8. 一名员工估计:同上,第 268-269 页;采访借调到 Sematech 的英特尔员工,2021 年;Larry D. Browning 和 Judy C. Shetler, Sematech:拯救美国半导体行业(德克萨斯 A&M 出版社,2000 年)。
  9. Sematech 组织研讨会:采访借调到 Sematech 的英特尔员工,2021 年。
  10. “问题的一半”:罗伯特·诺伊斯,1989 年 11 月 8 日在国会委员会作证;Peter N. Dunn,“GCA:工业政策的一课”,固态技术36,第 2 期(1993 年 12 月);Young, Silicon Sumo,第 270-276 页。
  11. “你完成了”: 2021 年采访彼得西蒙尼。
  12. 他决定当天购买:《采访彼得西蒙尼》,2021。
  13. “他们领先于时代”: 2021 年对甄子丹的采访;2021 年采访彼得·西蒙尼;Young, Silicon Sumo,第 262、285 页。
  14. 从尼康转而效忠: Young, Silicon Sumo,p。286.
  15. 无能为力:柏林,微芯片背后的人,p。304; Young,硅相扑,第 294-295 页;Jonathan Weber,“芯片制造先锋 GCA 公司关闭工厂:技术:6000 万美元的政府资金未能使马萨诸塞州公司恢复财务健康”,《洛杉矶时报》,1993 年 5 月 22 日。

第20章可以说不的日本

CHAPTER 20 The Japan That Can Say No

  1. Akio Morita 开始发现: Morita, Made in Japan , pp. 73, 110−120, 134。
  2. 一天十顿饭:内森,索尼,p。73.
  3. 日本的系统运行得更好:森田,日本制造,第 193、199、205 页。
  4. 太阳的季节 Ann Sherif,“速度的美学与政治的不合逻辑:石原慎太郎的文学首演”,日本论坛17,第 2 期(2005 年):185−211。
  5. “在经济上我们可以克服”:沃尔夫,日本的阴谋,p。16.
  6. “日本已成为一个非常重要的国家”: Akio Morita 和 Shintaro Ishihara, The Japan That Can Say No(Konbusha Publishing Ltd.,1996 年)。
  7. “丹麦的巨型版本”: Samuel Huntington,“Why International Primacy Matters”,国际安全(2009 年 1 月):75-76。
  8. “疯了”: Steven L. Herman,“国会中日本图书热门项目的盗版翻译”,美联社,1989 年 11 月 11 日。
  9. “我觉得美国读者不理解”:詹姆斯·弗拉尼根,“索尼首席执行​​官的美国抨击书让他失去了可信度”,洛杉矶时报,1989 年 10 月 11 日。
  10. “高科技是外交政策”: Harold Brown,“美国和日本:高科技是外交政策”, SAIS 评论9,第 2 期(1989 年秋季)。
  11. 中央情报局的任务:中央情报局,“1990 年代东亚的经济潜力:一篇投机论文”,CREST 数据库,1987 年。

第21章薯片王

CHAPTER 21 The Potato Chip King

  1. “先生。Spud”: 2021 年对美光员工的采访;乔治安德斯,“在马铃薯帝国,继承人剥离了多年的传统”,华尔街日报,2004 年 10 月 7 日;Laurence Zuckerman,“从 Spud 先生到 Chips 先生;美光背后的马铃薯大亨”,《纽约时报》,1996 年 2 月 8 日;Andrew E. Serwer,“Simplot 传奇:美国的炸薯条之王如何在半导体领域创造数十亿美元”,《财富》,2012 年 2 月 12 日。
  2. 所以他们求助于斯普德先生:采访沃德·帕金森,2021 年;Luc Olivier Bauer 和 E. Marshall Wilder,微芯片革命(独立出版,2020 年),第 279-280 页。
  3. 当地油腻的勺子: 2021 年采访 Elmer 的工作人员;2021 年采访沃德·帕金森。
  4. 投入数百万美元: Donald Woutat,“Maverick 芯片制造商转变立场:美光在发起价格战后支持保护主义”,《洛杉矶时报》,1985 年 12 月 16 日;Peter Burrows,“美光的复出小子”,《商业周刊》 ,1997 年 6 月 14 日。
  5. 损失和裁员: David E. Sanger,“美国芯片制造商的前景黯淡”,纽约时报,1985 年 10 月 29 日。
  6. 他们接受了他们在爱达荷州的局外人形象: David Staats,“高管的吹风机如何拯救记忆芯片——美光 40 年的故事”,爱达荷州政治家,2021 年 7 月 21 日。
  7. “弄巧成拙的策略”: Woutat,“Maverick Chip Maker 转变立场”。
  8. “法律规定他们不能那样做”: David E. Sanger,“发现日本芯片‘倾销’”,纽约时报,1985 年 8 月 3 日。
  9. “迄今为止生产成本最低”: 2021 年对 Ward Parkinson、Brian Shirley 和 Mark Durcan 的采访;Woutat,“Maverick 芯片制造商转变立场”。
  10. “以前没有写在论文中”:采访布赖恩雪莉和马克杜肯,2021;Yoshitaka Okada,“日本半导体产业的衰落”,日本半导体产业的发展(2006 年 1 月):41;Bauer 和 Wilder,《微芯片革命》,第 301-302 页。
  11. 为其余人员减薪: Bauer 和 Wilder,《微芯片革命》,第 286、302 页。
  12. “内存芯片是一项残酷、残酷的业务”:对 Mark Durcan、Ward Parkinson 和 Brian Shirley 的采访,2021 年。

第 22 章扰乱英特尔

CHAPTER 22 Disrupting Intel

  1. “我是个大忙人”: James Allworth,“英特尔的颠覆现已完成”, Medium,2020 年 11 月 11 日, https://jamesallworth.medium.com/intels-disruption-is-now-complete-d4fa771f0f2c
  2. “打屁股的匈牙利人”: Craig R. Barrett,Arnold Thackray 和 David C. Brock 于 2005 年 12 月 14 日和 2006 年 3 月 23 日在加利福尼亚州圣克拉拉接受采访(费城:化学遗产基金会,口述历史记录 0324)。
  3. 担心他错过了新闻: Andrew S. Grove,只有偏执狂才能生存:如何利用挑战每家公司的危机点(货币出版社,1999 年),第 117-118 页。
  4. 就像摩天轮上的一间小屋:格罗夫,只有偏执狂才能生存,第 88-90 页;Robert A. Burgelman,“Fading Memories:A Process Theory of Strategic Business Exist in Dynamic Environments”,《行政科学季刊》第 39 期,第 1 期(1994 年 3 月):41。
  5. 英特尔赢得了与 IBM 的一份小合同: Gerry Parker,“Intel's IBM PC Design Win”, Gerry Parker 的 Word Press 博客,2014 年 7 月 20 日, https://gerrythetravelhund.wordpress.com/tag/ibm-pc/;Jimmy Maher,“IBM PC 的完整历史,第一部分:世纪交易”, ars TECHNICA,2017 年 6 月 30 日, https://arstechnica.com/gadgets/2017/06/ibm-pc-history-第 1 部分/
  6. IBM 宣布推出其个人电脑: “IBM PC 的诞生”,IBM Debut Reference Room, https://www.ibm.com/ibm/history/exhibits/pc25/pc25_birth.html;“IBM 个人电脑发布会”,华尔道夫酒店,2019 年 1 月 23 日。
  7. 令人难以置信: Craig R. Barrett,2005 年 12 月 14 日和 2006 年 3 月 23 日在加利福尼亚州圣克拉拉接受 Arnold Thackray 和 David C. Brock 的采访。
  8. “争吵和争论”:格鲁夫,只有偏执狂才能生存,第 88-92 页。
  9. “建设性对抗”: Elizabeth Corcoran,“Intel CEO Andy Grove Steps Aside”,华盛顿邮报,1998 年 3 月 27 日;2021 年采访前英特尔员工。
  10. “控制部分”: Christophe Lecuyer,“面对日本挑战:英特尔制造业的复兴”,《商业历史评论》(2019 年 7 月);柏林,微芯片背后的人,p。180。
  11. “这就是你应该做的事情”: Lecuyer,“面对日本的挑战”,第 363-364 页;Craig R. Barrett,Arnold Thackray 和 David C. Brock 于 2005 年 12 月 14 日和 2006 年 3 月 23 日在加利福尼亚州圣克拉拉接受采访。Richard S. Tedlow, Andy Grove:美国商业偶像的生活和时代(企鹅, 2007),第 203.
  12. 更像一台经过微调的机器: Lecuyer,“面对日本挑战”,第 363、364、369、370 页;Craig R. Barrett,Arnold Thackray 和 David C. Brock 在加利福尼亚州圣克拉拉的采访,2005 年 12 月 14 日和 2006 年 3 月 23 日。第 65、79 页。
  13. 销量超过 IBM 自身: Therese Poletti,“Crucial Mistakes: IBM's Stumbles Open Door for Microsoft, Intel”,芝加哥论坛报,2001 年 8 月 13 日。

第23章“我的敌人的敌人”:韩国的崛起

CHAPTER 23 “My Enemy’s Enemy”: The Rise of Korea

  1. “大、强、永恒”: Geoffrey Cain,三星崛起(货币出版社,2020 年),第 10 页。33.
  2. 驾驶它在被占领的首都附近行驶:该隐,三星崛起,第 33-41 页。
  3. “通过商业服务国家”: Dong-Sung Cho 和 John A. Mathews, Tiger Technology(剑桥大学出版社,2007 年),第 105-106 页;该隐,三星崛起,第 40、41、46 页;关于李的财富,“韩国财阀在收入和股价方面的半个世纪的兴衰”,Yohap News Agency,2006 年 11 月 7 日, https: //www.yna.co.kr/view/AKR20110708154800008 。
  4. 努力赚钱: Si-on Park, Like Lee Byung-chul , p. 71; Cho 和 Mathews, Tiger Technology,p。112; Daniel Nenni 和 Don Dingee, Mobile Unleashed(Semi Wiki,2015 年);Kim Dong-Won 和 Stuart W. Leslie,“赢得市场还是赢得诺贝尔奖?KAIST 和晚期工业化的挑战,” Osiris 13 (1998): 167−170; Donald L. Benedict、KunMo Chung、Franklin A. Long、Thomas L. Martin 和 Frederick E. Terman,“关于设立韩国高等科学研究所的调查报告”,为美国国际开发署准备,1970 年 12 月, http://large.stanford.edu/history/kaist/docs/terman/summary/. 关于三星早期的困境,看韩泰半导体;例如。三星新闻编辑室,“半导体将是我最后的业务”,三星,2010 年 3 月 30 日,https://news.samsung.com/kr/91
  5. “不能仅仅通过观察来复制”: Park Si-on, Like Lee Byung-chul , pp. 399, 436。
  6. 至少 1 亿美元: Myung Oh 和 James F. Larson,《韩国的数字发展:建立信息社会》(Routledge,2011 年),p。54; Park Si-on, Like Lee Byung-chul , p. 386; Cho 和 Mathews, Tiger Technology,第 105、119、125 页;Lee Jae-goo,“我们为什么要从事半导体行业”,tr。Soyoung Oh, ZDNET Korea,1983 年 3 月 15 日, https: //zdnet.co.kr/view/?no=20110328005714 。
  7. 结果将是“致命的”: Tedlow, Andy Grove,p。218; Robert W. Crandall 和 Kenneth Flamm,改变规则(布鲁金斯学会出版社,1989 年),p。315; Susan Chira,“韩国的芯片制造商竞相赶超”,纽约时报,1985 年 7 月 15 日;“公司新闻:英特尔芯片协议”,《纽约时报》,1987 年 6 月 26 日。
  8. 贸易紧张局势也帮助了韩国公司: Richard E. Baldwin,“1986 年美日半导体协议的影响”,日本与世界经济,第 6 期,第 2 期(1994 年 6 月):136-137;Douglas A. Irwin,“Trade Policies and the Semiconductor Industry”,Anne O. Krueger 编辑,美国贸易政策的政治经济学(芝加哥大学出版社,1994 年),第 46-47 页。
  9. “我的敌人的敌人就是我的朋友”: Linsu Kim,“模仿创新:韩国技术学习的动力”,哥伦比亚大学东亚中心,1997 年,第 3 页。89,引用了 Zyrtek 以 210 万美元的费用转让先进生产知识的例子;2021 年采访沃德·帕金森;安德鲁波拉克,“美韩芯片关系发展”,纽约时报,1985 年 7 月 15 日。

第24章“这就是未来”

CHAPTER 24 “This Is the Future”

  1. Faggin 创造了一个芯片: Federico Faggin,“The Making of the First Microprocessor”,IEEE,2009;Federico Faggin,(Waterline,2021 年),特别是。通道。3.
  2. 对这个困境感到困惑: B. Hoeneisen 和 CA Mead,“微电子学的基本限制——I. MOS 技术,”固态电子学15,第 7 期(1972 年 7 月), https://authors.library.caltech.edu/54798/
  3. 一位才华横溢的计算机科学家: 2021 年采访 Lynn Conway,她想讨论约翰·加迪斯 (John Gaddis) 的细微差别,这让我感到惊讶,历史景观(牛津大学出版社,2004 年)。
  4. “隐形模式”: Dianne Lynch,“连线女性:工程师 Lynn Conway 的秘密”,ABC 新闻,2006 年 1 月 7 日。
  5. 奇怪的倒退: 2021 年采访林恩·康威。
  6. “你自己写”: “Lambda Magazine Lights the Way for VLSI Design”,IEEE 硅谷历史视频,YouTube 视频,2015 年 7 月 27 日,00:01:40, https://www.youtube.com/watch? v=DEYbQiXvbnc ; “VL​​SI 的历史 – C. Mead – 2/1/2011,”加州理工学院,YouTube 视频,2018 年 5 月 29 日, https://www.youtube.com/watch?v= okZBhJ-KvaY 。
  7. 古腾堡时刻已经到来: “1981 年电子成就奖”,密歇根大学, https://ai.eecs.umich.edu/people/conway/Awards/Electronics/ElectAchiev.html;2021 年采访 Lynn Conway 和 Carver Mead。
  8. 芯片设计人员供应充足: Van Atta 等人,DARPADARPA项目 II的历史回顾,1990 年 2 月,AD-A239 925,p. 17-5。
  9. 帮助公司和教授保持摩尔定律的活力: 2021 年 Paul Losleben 访谈;Van Atta 等人,DARPA技术成就,p。17-1。
  10. 由校友创立和建造:采访 David Hodges、Steve Director、Aart de Geus、Alberto Sangiovanni-Vincentelli 和 Rob Rutenbar;“1984 年年度报告”,半导体研究公司,1984 年, https://www.src.org/src/story/timeline
  11. 如何存储和传达信息: David Morton 对 Irwin Jacobs 的采访,IEEE 历史中心,1999 年 10 月 29 日。
  12. “这就是未来”: Daniel J. Costello, Jr. 和 David Forney, Jr.,“通道编码:通道容量之路”,IEEE 95 会议记录,第 6 期(2007 年 6 月);O. Aftab、P. Cheung、A. Kim、S. Thakkar 和 N. Yeddanapudi,“信息理论和数字时代”,6.933 项目历史,麻省理工学院, https: //web.mit.edu/6.933/www/ 2001 年秋季/Shannon2.pdf;1995 年 5 月 10 日,电气工程史中心的 Andrew Goldstein 对 David Forney Jr. 的采访;Daniel Nenni,“高通的详细历史”, SemiWiki,2018 年 3 月 19 日, https: //semiwiki.com/general/7353-a-detailed-history-of-qualcomm/ 。

第 25 章克格勃总局 T

CHAPTER 25 The KGB’s Directorate T

  1. 留在莫斯科喝醉:维特罗夫的生活细节很大程度上来自谢尔盖·科斯汀和埃里克·雷诺,告别:二十世纪最伟大的间谍故事(亚马逊穿越,2011 年)。
  2. “提高其生产集成电路的能力”:中央情报局,“苏联情报部门的技术获取努力”,1982 年 6 月 18 日,第 3 页。15、 https://www.cia.gov/readingroom/docs/DOC_0000261337.pdf;Philip Hanson,苏联工业间谍(皇家国际事务研究所,1987 年)。
  3. “跌倒”后死亡: Sergey Chertoprud, Naucho-Tekhnicheskaia Razvedka(Olma Press,2002),p. 283; 丹妮拉·艾科诺(Daniela Iacono),“一位陷入死亡的英国银行家”,联合新闻国际,1984 年 5 月 15 日;Michael S. Malone,“进入硅谷地下”,纽约时报,1982 年 5 月 30 日。
  4. 关闭苏联研究单位: Jay Tuck, High-Tech Espionage (St. Martin's Press, 1986), p. 107; 西蒙诺夫, Nesostoyavshayasya Informatsionnaya Revolyutsiya,p。34.
  5. 通过空壳公司: Edgar Ulsamer,“莫斯科的技术寄生虫”,空军杂志,1984 年 12 月 1 日。
  6. 兴奋剂、封装和测试芯片:中央情报局,“苏联对具有军事意义的西方技术的收购:更新”,1985 年 9 月,第 25, http: //insidethecoldwar.org/sites/default/files/documents/CIA%20Report%20on%20Soviet%20Acquisition%20of%20Militarily%20Significant%20Western%20Technology%20September%201985.pdf
  7. 与法国情报部门有关: Kostin 和 Raynaud,告别
  8. 苏联偷了多少:汉森,苏联工业间谍;中央情报局,“苏联对具有重要军事意义的西方技术的收购:更新”;科斯汀和雷诺,告别;蒂埃里·沃尔顿(Thierry Wolton),《法国克格勃》(Club Express,1986 年)。
  9. 总是落后五年:中央情报局,“苏联计算机技术:追赶的前景渺茫”,国家安全档案馆,1985 年 3 月,第 4、 https://nsarchive.gwu.edu/document/22579-document-02-central-intelligence-agency-soviet;Bruce B. Weyrauch,“Operation Exodus”,《计算机/法律杂志》第 7 期,第 2 期(1986 年秋季);汉森,苏联工业间谍;Jon Zonderman,“监管高科技出口”,《纽约时报》,1983 年 11 月 27 日。

第26章“大规模杀伤性武器”:抵消的影响

CHAPTER 26 “Weapons of Mass Destruction”: The Impact of the Offset

  1. “大规模杀伤性武器”: Dale Roy Herspring,苏联最高指挥部,1967-1989 年(普林斯顿大学出版社,2016 年),第 10 页。175.
  2. “要求”: Christopher Andrew 和 Oleg Gordievsky,“1983 年 KAL 航班的击落表明苏联缺乏虚构 007 的技能”,《洛杉矶时报》,1990 年 11 月 11 日。
  3. 具有明显优势: Brian A Davenport,“The Ogarkov Ouster”,战略研究杂志14,第 2 期(1991 年):133;中央情报局和国防部,“美国和苏联战略力量:联合网络评估”,国防部长,1983 年 11 月 14 日, https://nsarchive2.gwu.edu/NSAEBB/NSAEBB428/docs/1.US%20and%20Soviet% 20Strategic%20Forces%20Joint%20Net%20Assessment.pdf
  4. “军事技术革命”:海军分析中心,奥加科夫元帅谈现代战争:1977−1985年,AD-A176 138,p。27; Dima P. Adamsky,“镜花水月:苏联军事技术革命和美国军事革命”,《战略研究杂志》,第 31 期,第 2 期(2008 年)。
  5. 佩里的“抵消战略”正在奏效:大卫·伯巴赫、布伦丹·里滕豪斯·格林和本杰明·弗里德曼对抵消技术的出色概述,所有这些技术都从根本上依赖于半导体,“军事革命的技术”, Harvey Sapolsky、Benjamin Friedman 和 Brendan Green 编辑,《冷战以来的美国军事创新:无破坏的创造》(Routledge,2012 年),第 14-42 页;中央情报局,“苏联国防工业:应对军事技术挑战”,中央情报局历史回顾计划,1987 年 7 月,p。17、 https://www.cia.gov/readingroom/docs/DOC_0000499526.pdf;亚当斯基,“透过镜子”,p。260。
  6. 苏联第一台使用集成电路的导弹制导计算机: Anatoly Krivonosov,“Khartron:火箭制导系统计算机”,Boris Malinovsky,“乌克兰计算机科学与技术史”,tr。Slava Gerovitch,苏联太空计划中的计算,2002 年 12 月 16 日, https ://web.mit.edu/slava/space/essays/essay-krivonosov.htm ;Donald MacKenzie,“苏联与战略导弹制导”,国际安全13,第 2 期(1988 年秋季);Slava Gerovitch 对 Georgii Priss 的采访,《苏联太空计划中的计算》,2002 年 5 月 23 日, https: //web.mit.edu/slava/space/interview/interview-priss.htm#q3 。
  7. 计算了他们自己的目标路径: MacKenzie,“苏联和战略导弹制导”,第 30-32、35 页。
  8. 摧毁了 98% 的苏联洲际弹道导弹: MacKenzie,“苏联与战略导弹制导”,第 10 页。52,引用 CEP 为 0.06 海里;Pavel Podvig,“不存在的机会之窗:1970 年代的苏联军事集结”,国际安全(2008 年夏季):129,引用了 0.35-0.43 公里的 CEP。还有其他一些可以比较导弹的变量,包括它们携带的弹头的大小和数量,以及它们可以发射或重新瞄准的速度。但美国的准确性优势的基本趋势是成立的;98% 的数据来自 John G. Hines、Ellis M. Mishulovich 和 John F. Shull,《苏联意图》,1965-1985 年,卷。2 (BDM Federal, Inc., 1995), pp. 46, 90。请注意,这 98% 可能大大夸大了美国的能力,但仍然是苏联恐惧的证据。参照。Brendan R. Green 和 Austin Long,“不在场的 MAD:苏联对冷战后期核平衡的反应”,《安全研究》第 26 期,第 4 期(2017 年 7 月 7 日)。
  9. 追踪苏联潜艇: Owen R. Cote, Jr.,“第三次战役:美国海军与苏联潜艇无声的冷战斗争中的创新”,纽波特论文,海军战争学院,2003 年;Joel S. Wit,“反潜战的进展”,科学 American 244,第 2 期(1981 年 2 月):31-41;DL Slotnick,“并行处理器的概念和发展:个人回忆录” ,计算历史年鉴4,第 1 期(1982 年 1 月-3 月);Van Atta 等人,DARPA技术成就 II;Christopher A. Ford 和 David A. Rosenberg,“里根海上战略的海军情报基础”,《战略研究杂志》,第 28 期,第 2 期(2005 年 4 月):398;John G. Hines、Ellis M. Mishulovich 和 John F. Shull,《1965-1985 年苏联意图》,卷。1 (BDM Federal, Inc., 1995), p. 75; Green and Long,“The MAD Who Wasn't There”,第 607、639 页。在 1980 年代,苏联 SSBN 导弹的可靠性也存在重大问题;见史蒂文·J·扎洛加,克里姆林宫的核剑:1945-2000 年俄罗斯战略核力量的兴衰(史密森尼书籍,2014 年),第 10 页。188.
  10. “战略武器的劣势”: Green 和 Long,“The MAD Who Wasn't There”,第 3 页。617.
  11. 苏维埃国家的生存: Danilevich 引自 Hines、Mishulovich 和 Shull,《 1965-1985 年苏联意图》,第 1卷。1,第 57; Dale R. Herspring,“Nikolay Ogarkov 和苏联军事事务中的科技革命”,比较战略6,第 1 期(1987 年);Mary C. Fitzgerald,“苏联对未来战争的看法:新技术的影响”,《国防分析》第 7 期,第 2-3 期(1991 年)。苏联官员对指挥控制和通信系统的生存能力深表担忧;参见 Hines、Mishulovich 和 Shull, 1965-1985 年的苏联意图,卷。1,第 90; 瓦西里·彼得罗夫元帅在 1983 年引述北约计划“创造和利用‘解除武装’第一次[常规]罢工的潜力”,托马斯·M.尼科尔斯,神圣的事业:针对苏联国家的军民冲突安全性,1917-1992年(NCROL,1993 年),p。117; Mary C. Fitzgerald,“奥加科夫元帅谈现代战区行动”,海军战争学院评论39,第 4 期(1986 年);Mary C. Fitzgerald,“奥加科夫元帅和苏联军事事务中的新革命”,国防分析3,第 1 期(1987 年)。
  12. “更加自律”:米哈伊尔·戈尔巴乔夫,“ Zasedanie Politbyuro Tsk Kpss 30 Iiulia Goda ”,载于Sobranie Sochinenii,第 9 卷(Ves' Mir,2008 年),第 339-343 页。我在这里翻译得很自由。
  13. 奥索金被移除: 2021 年采访谢尔盖奥索金。
  14. 资本投资的八倍:西蒙诺夫, Nesostoyavshayasya Informatsionnaya Revolutsiya,p。70; Seymour Goodman 和 William K. McHenry,“苏联计算机工业:两个部门的故事”,ACM 通讯(1991 年 1 月):32。
  15. 缺乏国际供应链: VV Zhurkin,“ Ispolzovanie Ssha Noveishhikh Dostizhenii Nauki i Tekhniki v Sfere Vneshnei Politiki ”,科学院档案馆,1987 年 8 月 7 日。
  16. 东德芯片输出: Charles S. Maier, Dissolution(普林斯顿大学出版社,1999 年),第 74-75 页。

第27章战争英雄

CHAPTER 27 War Hero

  1. 诺曼·施华蔻:罗伯特·D·麦克法登,“Gen。H. Norman Schwarzkopf,美国海湾战争指挥官,享年 78 岁,”纽约时报,2012 年 12 月 27 日。
  2. 波斯湾战争已经开始: Rick Aktinson,十字军东征:波斯湾战争的不为人知的故事(Mariner Books,1994 年),第 35-37 页。
  3. 美国空袭试图斩首: “剧院开幕法案”,华盛顿邮报,1998;阿克丁森,十字军东征,p。37.
  4. 宝石路激光制导炸弹: 2021 年史蒂夫·罗默曼采访中有关宝石路电子设备的详细信息。
  5. 十三次: Stephen P. Rosen,“The Impact of the Office of Net Assessment on the American Military in the Matter of the Revolution of Military Affairs”,《战略研究杂志》,第 33 期,第 4 期(2010 年):480。
  6. “一万个美国人”: 2021 年采访史蒂夫·罗默曼。
  7. “高科技作品”: Bobby R. Inman、Joseph S. Nye Jr.、William J. Perry 和 Roger K. Smith,“海湾战争的教训”,华盛顿季刊15,第 1 期(1992 年):68 ; Benjamin S. Lambeth,沙漠风暴及其意义(兰德公司,1992 年)。
  8. “战争英雄地位”: William J. Broad,“海湾战争:高科技;计算机芯片可能成为战争英雄,”纽约时报,1991 年 1 月 21 日;Barry D. Watts,制导弹药和战斗网络的六个十年:进展和前景(战略和预算评估中心,2007 年),p。146; 采访史蒂夫·罗默曼。
  9. 伊拉克的迅速投降: Mary C. Fitzgerald,“苏联军队和海湾地区的新‘技术行动’”,海军战争学院评论44,第 4 期(1991 年秋季):16−43, https://www.jstor .org/stable/44638558;Stuart Kaufman,“1991 年海湾战争和军事学说的教训”,《斯拉夫军事研究杂志》第 6 期,第 3 期(1993 年);Graham E. Fuller,“莫斯科和海湾战争”,外交事务(1991 年夏季);Gilberto Villahermosa,“沙漠风暴:苏联观点”,外国军事研究办公室,2005 年 5 月 25 日,第 4.

第28章“冷战结束,你赢了”

CHAPTER 28 “The Cold War Is Over and You Have Won”

  1. 政府支持的过度投资: Michael Pettis, The Great Rebalancing(普林斯顿大学出版社,2013 年)。
  2. 三星削弱了日本的竞争对手: Yoshitaka Okada,“日本半导体工业的衰落”,载于 Yoshitaka Okada 编辑的Struggles for Survival (Springer, 2006), p. 72.
  3. “你晚上睡不着觉”: Marie Anchordoguy, Reprogramming Japan(康奈尔大学出版社,2005 年),第 3 页。192.
  4. 公司未能削减投资: Sumio Saruyama 和 Peng Xu,产能过剩与退出困难:来自日本电子行业的证据(Springer Singapore,2021);“决心推动了 CCD‘电眼’的发展,”索尼, https://www.sony.com/en/SonyInfo/CorporateInfo/History/SonyHistory/2-11.html
  5. 这种新型存储芯片: Kenji Hall,“Fujio Masuoka:感谢您的记忆”,彭博社,2006 年 4 月 3 日;Falan Yinung,“闪存市场的崛起:其对企业行为和全球半导体贸易模式的影响”,国际商业与经济杂志(2007 年 7 月)。
  6. 到 1998 年 20%: Andrew Pollack,“美国筹码的收益是日本的损失”,《纽约时报》,1991 年 1 月 3 日;冈田,“日本半导体产业的衰落”,第 10 页。41; “半导体行业的趋势”,日本半导体历史博物馆, https://www.shmj.or.jp/english/trends/trd90s.html
  7. 支持伊拉克的邻国:日本外务省,“海湾危机如何开始和结束”, 1991 年外交蓝皮书 https://www.mofa.go.jp/policy/other/bluebook/1991/1991-2-1 .htm ; 日本外务省,“日本对海湾危机的反应”, 1991 年外交蓝皮书 https://www.mofa.go.jp/policy/other/bluebook/1991/1991-2-2.htm Kent E. Calder,“美国、日本和海湾地区”,笹川和平基金会,2015 年 8 月,第 31; TR Reid,“日本的新挫折”,华盛顿邮报,1991 年 3 月 17 日。
  8. “我们不要争论谁赢了”: “G-Day:苏联总统戈尔巴乔夫访问斯坦福商学院”,斯坦福商学院,1990 年 9 月, https://www.gsb.stanford.edu/experience/news-history /历史/g-day-soviet-president-gorbachev-visits-stanford-business-school ; 大卫·雷姆尼克,“在美国,戈尔巴乔夫试图兜售梦想”,华盛顿邮报,1990 年 6 月 6 日。
  9. “从 5 岁开始就有电脑”: Gelb 在 1992 年首次讲述了这个故事;我引用了他 2011 年关于该主题的文章;Leslie H. Gelb,“外交事务;谁赢得了冷战?” 纽约时报,1992 年 8 月 20 日;Leslie H. Gelb,“被遗忘的冷战:20 年后,关于美国胜利的神话依然存在”,《每日野兽》 ,2017 年 7 月 14 日。
  10. 开心乐园餐玩具: 2021 年彼得·戈登访谈。

第 29 章“我们希望在台湾建立半导体产业”

CHAPTER 29 “We Want a Semiconductor Industry in Taiwan”

  1. “你需要多少钱”: Wang, KT Li and the Taiwan Experience , p. 217; Morris Chang 的口述历史,由 Alan Patterson 拍摄,2007 年 8 月 24 日,计算机历史博物馆。
  2. 公司能力落后: Tekla S. Perry,“Morris Chang:铸造之父”, IEEE Spectrum,2011 年 4 月 19 日;“斯坦福工程英雄讲座:Morris Chang 与总裁 John L. Hennessy 的对话”,斯坦福在线,YouTube 视频,2004 年 4 月 25 日,第 36 分钟左右, https://www.youtube.com/watch?v=wEh3ZgbvBrE
  3. 经济“战争”: “TI 董事会访问台湾 1978”,德州仪器特别收藏,90-69 TI 董事会访问台湾,DeGolyer 图书馆,南卫理公会大学。
  4. “放牧”: Morris Chang的口述历史,计算机历史博物馆。
  5. “对我来说陌生的地方”: “Morris Chang 的最后一次演讲”,tr。Kevin Xu,互联通讯,2021 年 9 月 12 日, https://interconnected.blog/morris-changs-last-speech;关于拒绝工作机会,请参阅 L. Sophia Wang 编辑, KT Li Oral History (2nd edition, 2001), pp. 239‒40,感谢 Mindy Tu 的翻译;“斯坦福工程英雄讲座:Morris Chang 与总裁 John L. Hennessy 的对话”,大约在第 34 分钟, https://www.youtube.com/watch?v=wEh3ZgbvBrE。关于常的德州身份:2022 年专访莫里斯·常。
  6. 余地:张忠谋口述历史,计算机历史博物馆。
  7. “丰富的应用”: “1976 Morris Chang Planning Doc”,德州仪器特别收藏,Fred Bucy 论文,DeGolyer 图书馆,南卫理公会大学。
  8. 已经在台湾渗透:林凌飞对 Chintay Shih 的采访,计算机历史博物馆,2011 年 2 月 21 日;国家研究委员会,“附录 A3:台湾工业技术研究所”, 21 世纪制造业(国家科学院出版社,2013 年);莫里斯·张口述历史,计算机历史博物馆。
  9. “这不是其中之一”: Douglas B. Fuller,“国家建设的全球化:台湾高科技产品的产业政策”,工作文件,麻省理工学院,2002 年。
  10. 27.5% 的股份: Rene Raaijmakers, ASML 的建筑师(Techwatch Books,2018 年),ch。57. 关于飞利浦的知识产权转让,参见 John A. Mathews,“A Silicon Valley of the East”, California Management Review (1997):36;Daniel Nenni,“台积电简史”, SemiWiki,2012 年 8 月 2 日。
  11. 台湾国家项目: “斯坦福工程英雄讲座:张忠谋与总裁约翰·L·轩尼诗对话”;斯坦福大学图书馆 Rob Walker 对 Donald Brooks 的采访,2000 年 2 月 8 日,1:45, https: //exhibits.stanford.edu/silicongenesis/catalog/cj789gh7170 。
  12. 与美国芯片行业的深厚联系: “TSMC Announces Resignation of Don Brooks”, EE Times,1997 年 3 月 7 日;唐纳德·布鲁克斯接受罗伯·沃克的采访,1:44;“1995 年年度报告”,台积电,1995 年;关于教育联系,请参阅 Douglas B. Fuller,“台湾产业政策的日益无关紧要,2016-2020 年”,Gunter Schubert 和 Chun-Yi Lee 编辑,台湾蔡英文第一届政府期间:风雨飘摇沃特斯(Routledge,2020 年),p。15.
  13. 台湾和硅谷受益: AnnaLee Saxenian,区域优势:硅谷和 128 号公路的文化与竞争(哈佛大学出版社,1994 年);AnnaLee Saxenian, The New Argonauts:全球经济中的区域优势(哈佛大学出版社,2006 年)。

第 30 章“所有人都必须制造半导体”

CHAPTER 30 “All People Must Make Semiconductors”

  1. 相同的存储容量: Jonathan Pollack,“转型中的中国电子工业”,兰德公司,N-2306,1985 年 5 月;David Dorman,“中国经济改革中的军事当务之急:电子政治,1949−1999 年”,博士论文,马里兰大学帕克分校,2002 年;在 1KB DRAM 上,请参见 Richard Baum,“DOS ex Machina”,Denis Fred Simon 和 Merle Goldman 编着的《后毛泽东时代中国的科学与技术》(哈佛大学亚洲中心,1989 年),第 3 页。357.
  2. 中国工程师打造了他们的第一块集成电路:朱一伟,《中国集成电路产业发展论文集》,tr。Zoe Huang (2006),第 140-144 页。
  3. “经过几年的研究”:国家研究委员会,“中华人民共和国的固态物理学:美国固态物理学代表团的旅行报告”,1976 年,第 3 页。89.
  4. 《惊天动地的群众运动》:《上海工人大力发展电子工业》,1969 年 10 月 9 日,《人民日报》译文《中国大陆报》第4520 期,1969 年 10 月 21 日,第 11-13 页。
  5. 中国的社会主义乌托邦: Denis Fred Simon 和 Detlef Rehn,《中国的技术创新:上海半导体产业案例》(Ballinger Publishing Company,1988 年),第 47、50 页;Lowell Dittmer,“死亡与变形”,亚洲研究杂志40,第 3 期(1981 年 5 月):463。
  6. “人人都要做半导体”:兰友航,《中国商用电子显微镜的建设》,影像与电子物理进展, 96(1996):821;Sungho Rho、Keun Lee 和 Seong Hee Kim,“中国半导体行业的有限追赶:从行业创新系统的角度看”, Millennial Asia(2015 年 8 月 19 日):159。
  7. “浪费了这么多”:华瓜峰,1975 年 9 月 26 日,引自 Roderick MacFarquhar 和 Michael Schoenhals,毛泽东的最后革命(Belknap Press,2008 年),第 400-401 页。
  8. “自我荣耀”:国家研究委员会,“中华人民共和国的固体物理学”,第 10 页。151.
  9. 政治思想家: Hoddeson 和 Daitch, True Genius,p。277.
  10. 1500 台计算机: Baum,“DOS ex Machina”,第 347-348 页;国家研究委员会,“中华人民共和国的固态物理学”,第 52-53 页。
  11. 武器系统、消费电子产品和计算机: Simon 和 Rehn,《中国的技术创新》,第 15、59、66 页;Baum,“DOS ex Machina”,第 347-348 页。
  12. “‘第三台出口机器’”: Simon 和 Rehn,中国的技术创新,第 17、27、48 页。

第 31 章“与中国人分享上帝的爱”

CHAPTER 31 “Sharing God’s Love with the Chinese”

  1. “与中国人分享上帝的爱”: Evelyn Iritani,“中国的下一个挑战:掌握微芯片”,洛杉矶时报,2002 年 10 月 22 日。
  2. 包括一座教堂:安德鲁·罗斯,《快船到中国》(Vintage Books,2007 年),第 10 页。250。
  3. 到 2010 年达到 13%: Antonio Varas、Raj Varadarajan、Jimmy Goodrich 和 Falan Yinug,“政府激励措施和美国在半导体制造领域的竞争力”,波士顿咨询集团和半导体行业协会(2020 年 9 月),第 10 页。7.
  4. 特许半导体公司: John A. Matthews,“东方硅谷”,加州管理评论(1997 年)。
  5. 像一场小鸡游戏: 2021 年三星高管访谈。
  6. 三星有资本继续投资:关于信贷补贴,参见 S. Ran Kim,“The Korean System of Innovation and the Semiconductor Industry”, Industrial and Corporate Change 7,第 2 期(1998 年 6 月 1 日):297−298。
  7. 填充行李箱: 2021 年中国技术分析师访谈。
  8. 运行 TI 在全球的工厂: Peter Clarke,“ST 工艺技术是 Chang 的下一个中国铸造厂的基础”,tr。克劳斯宋, EE 新闻模拟2020 年 2 月 24 日;《商业人物周刊:中国半导体之父——理查德·张》,中央电视台,YouTube 视频,2010 年 4 月 29 日,https://www.youtube.com/watch?v=NVHAyrGRM2Ehttp://magazine.sina.com/bg/southernpeopleweekly/2009045/2009-12-09/ba80442.htmlhttps://www.coolloud.org.tw/node/6695
  9. 大多数早期结果: Douglas B. Fuller, Paper Tigers,Hidden Dragons(牛津大学出版社,2016 年),第 122-126 页;John VerWey,“中国半导体产业政策:过去与现在”,美国国际贸易委员会国际商业与经济期刊(2019 年 7 月):11。
  10. 甜蜜的金融交易:这是中国芯片行业领先专家 Doug Fuller 在《纸老虎,藏龙》第 10 页中的判断。122.
  11. “位于中国的晶圆厂”: Fuller, Paper Tigers, Hidden Dragons , p. 125; 李印,“从经典失败到全球竞争对手:中国半导体行业的商业组织和经济发展”,硕士论文,马萨诸塞大学洛厄尔分校,第 32-33 页。
  12. 台湾塑料王朝: Lee Chen Yee 和 David Lin,“Hua Hong NEC,Grace Close to Merger”,路透社,2011 年 12 月 1 日。
  13. Neil Bush: “报告称,中国上海格雷斯半导体在新工厂破土动工”, EE Times,2000 年 11 月 20 日;Warren Vieth 和 Lianne Hart,“布什的兄弟有合同帮助中国芯片制造商”,《洛杉矶时报》,2003 年 11 月 27 日。
  14. 难以获得客户: Ming-chin Monique Chu,东亚计算机芯片大战(Routledge,2013 年),第 212-213 页;“江泽民长子江绵恒的快速成功,多年来一直受到中国学术界的质疑”,南华早报,2015 年 1 月 9 日。5个;Michael S. Chase、Kevin L. Pollpeter 和 James C. Mulvenon,“上海:海峡两岸信息技术和投资流动的经济和政治影响(技术报告)”,兰德公司,2004 年 7 月 26 日,第 127-135 页。
  15. 高盛、摩托罗拉和东芝: “Richard Chang:台湾的硅入侵”,彭博商业周刊,2002 年 12 月 9 日;罗斯,《快船到中国》,p。250。
  16. 中芯国际一半的启动资金: Chase et al., “Shanghaied”, p. 149.
  17. 台湾四百人: “Richard Chang 和他的中芯国际团队”, Cheers Magazine,2000 年 4 月 1 日, https: //www.cheers.com.tw/article/article.action?id=5053843 。
  18. 受过外国培训的劳动力: Fuller, Paper Tigers, Hidden Dragons , pp. 132, 134−135;VerWey,“中国半导体产业政策”,第 11-12 页;李印,“从经典失败到全球竞争对手”,第 45-48 页;卢尔浩,《中国半导体产业发展模式,2000-2005》(中国版道提chanye fazhan moshi),博士论文,国立政治大学,台北,台湾,2008,第33-35页,感谢Claus Soong翻译;罗斯,《快船到中国》,p。248.
  19. 降低销售税:陈寅寅,“上海半导体产业发展的政治经济学,1956-2006”,论文,国立台湾大学,2007,第 71-72 页;卢,中国的发展模式 半导体工业,第 75-77 页。感谢克劳斯宋翻译这些资料。
  20. 接近前沿: Yin Li,“从经典失败到全球竞争对手”,第 45-48 页。
  21. 有望成为一流的铸造厂: Fuller, Paper Tigers, Hidden Dragons , pp. 132, 136; “半导体制造国际公司宣布拟在联交所和纽约证券交易所双重上市”,中芯国际,2004 年 3 月 7 日, https://www.smics.com/en/site/news_read/4212;“芯片制造商中芯国际首次亮相,”CNN,2004 年 3 月 18 日。

第32章光刻战争

CHAPTER 32 Lithography Wars

  1. 他给了卡拉瑟斯 2 亿美元:采访约翰·卡拉瑟斯,2021;本章受益于对 Vivek Bakshi、Chris Mack、Chuck Gwyn、David Attwood、Frits van Houts、John Taylor、John Carruthers、Bill Siegle、Stefan Wurm、Tony Yen、Shang-yi Chiang 以及其他要求不要被点名,没有人对结论负责。
  2. “光刻战争”: Mark L. Schattenburg,“‘三光束’会议的历史、信息的诞生和光刻战争时代”, https://eipbn.org/2020/wp-content/uploads/2015 /01/EIPBN_history.pdf
  3. 自愿加入: Peter Van Den Hurk,“告别‘顶级人物大家庭’”,ASML,2021 年 4 月 23 日, https://www.asml.com/en/news/stories/2021/frits-van -hout-retires-from-asml
  4. “没有设施,也没有钱”: 2021 年采访 Frits van Hout。
  5. 日本的竞争对手: Rene Raiijmakers,“技术所有权不是与生俱来的权利”, Bits & Chips,2021 年 6 月 24 日。
  6. 建立伙伴关系: Fritz van Hout 访谈,2021 年;《光刻大战(中):台积电的火是如何拯救光刻巨头ASML的?》iNews,2022 年 2 月 5 日, https: //inf.news/en/news/5620365e89323be681610733c6a32d22.html 。
  7. “单极时刻”: Charles Krauthammer,“单极时刻”,外交事务,1990 年 9 月 18 日。
  8. “无边界世界”: Kenichi Ohmae,“管理无边界世界”,哈佛商业评论(1989 年 5 月至 6 月)。
  9. 每年都盈利:根据彭博数据。
  10. “95% 的大猩猩”:采访约翰·泰勒,2021 年。
  11. ASML 是唯一剩下的光刻公司: Chuck Gwyn 和 Stefan Wurm,“EUV LLC:历史视角”,Bakshi 编辑,EUV光刻(SPIE,2008 年);2021 年采访约翰·卡拉瑟斯和约翰·泰勒。
  12. 华盛顿几乎没有人担心: 2021 年对 Kenneth Flamm 和 Richard Van Atta 的采访。
  13. 工作,而不是地缘政治: David Lammers,“美国在 EUV 上对 ASML 表示满意”, EE Times,1999 年 2 月 24 日;这份媒体报道引用了与美国政府的一项协议,据此 ASML 承诺在美国生产其部分机器或 ASML,尽管多位前官员表示,这笔交易听起来很合理,而且它可能是非正式的而不是正式的。今天,ASML 在康涅狄格州的一家制造工厂生产每个 EUV 工具的一部分,因此如果事实上它做出了这样的承诺,它似乎会支持它的交易。
  14. 政府决定让这一安排继续进行:我的采访对象中没有一个人认为外交政策因素对这一决定至关重要,许多人说他们不记得有任何关于这个话题的讨论。
  15. “美国政府的所有 EUV 技术”: Don Clark 和 Glenn Simpson,“反对 SVG 出售荷兰人担心外国竞争”,华尔街日报,2001 年 4 月 26 日;采访光刻行业专家,2021;2021 年对迪克·范·阿塔的采访;2021 年采访前商务部官员。
  16. EUV 没有上榜: Clark 和 Simpson,“SVG 销售的反对者让荷兰人担心外国竞争。”
  17. 美国、日本、斯洛文尼亚和希腊:约翰·泰勒访谈,2021 年。

第33章创新者的困境

CHAPTER 33 The Innovator’s Dilemma

  1. “英特尔准备就绪”: “第一台英特尔 Mac(2006 年 1 月 10 日)”,关于 Steve Jobs.com,YouTube 视频,2009 年 9 月 18 日, https://www.youtube.com/watch?v= cp49Tmmtmf8 。
  2. 更经常打领带: 2021 年对资深英特尔高管的采访。
  3. 挤占英特尔在 x86 芯片上事实上的垄断地位: Alexis C. Madrigal,“Paul Otellini 的英特尔:构建未来的公司能否生存下来?” 大西洋,2013 年 5 月 16 日;2021 年对四位前英特尔高管的采访。
  4. 保卫城堡的护城河是 x86:迈克尔·布鲁克访谈,2021 年。
  5. 赢得了近乎垄断:库尔特舒勒,“半导体放缓?投资!” 半导体工程,2012 年 1 月 26 日。
  6. “我们唯一的机会”: Robin Saxby 访谈,2021 年;“Robin Saxby 爵士:ARM 架构是在 Acorn 计算机中发明的”,Anu Partha,YouTube 视频,2017 年 6 月 1 日, https://www.youtube.com/watch?v=jxUT3wE5Kwg;Don Dingee 和 Daniel Nenni, Mobile Unleashed:我们设备中 ARM 处理器的起源和演变(SemiWiki LLC,2015),特别是。页。42; “校友获得电气和电子工程师协会 (IEEE) 颁发的最高荣誉,”利物浦大学,2019 年 5 月 17 日。
  7. Zoom-esque 视频会议: 2021 年对英特尔前高管的采访。
  8. 移动设备似乎是一场疯狂的赌博: 2020 年采访 Ted Odell 和 2021 年 Will Swope。
  9. “体积是所有人想象的 100 倍”: Alexis C. Madrigal,“Paul Otellini 的英特尔”。
  10. 英特尔从未找到在移动领域赢得立足点的方法: Joel Hruska,“英特尔如何失去移动市场,第 2 部分:Atom 的崛起与忽视”, Extreme Tech,2020 年 12 月 3 日;Joel Hruska,“英特尔如何损失 100 亿美元和移动市场”, Extreme Tech,2020 年 12 月 3 日;Mark Lipacis 等人,“半导体:计算领域的第四次构造转变:到并行处理/物联网模型,”杰弗里斯 研究笔记,2017 年 7 月 10 日;与 Michael Bruck 和 Will Swope 的对话有助于明确这一点;Varas 等人,“在不确定的领域加强全球半导体供应链”。
  11. “不想受到利润冲击”: 2021 年对英特尔前高管的采访。

第34章跑得更快?

CHAPTER 34 Running Faster?

  1. “为胜利而战”: Andy Grove,“Andy Grove:美国如何创造就业机会”,《商业周刊》 ,2010 年 7 月 1 日。
  2. “我怀疑他们永远不会赶上”:同上。
  3. 每个芯片上有十亿个晶体管: Jon Stokes,“二十亿晶体管野兽:POWER7 和 Niagara 3”, Ars Technica,2010 年 2 月 8 日。
  4. 控制了大约四分之三的市场: Wally Rhines,“电子设计自动化行业的竞争动态”, SemiWiki,2019 年 8 月 23 日。
  5. 7.3 级地震: Mark Veverka,“台湾地震发出警钟,但影响可能是短暂的”,巴伦周刊,1999 年 9 月 27 日。
  6. 第五次花费的时间更长: Jonathan Moore,“Fast Chips, Faster Cleanup”,《商业周刊》 ,1999 年 10 月 11 日。
  7. 中断是有限的: Baker Li,道琼斯通讯社,“台湾地震后零部件短缺似乎正在消退”,华尔街日报,1999 年 11 月 9 日。
  8. 5 自 1900 年: 2021 年采访无晶圆公司高管;“世界上最大的 20 次地震”,美国地质勘探局, https: //www.usgs.gov/natural-hazards/earthquake-hazards/science/20-largest-earthquakes-world?qt-science_center_objects=0#qt-science_center_objects 。
  9. “负责任的利益相关者”: Robert Zoellick 演讲,2005 年 9 月 21 日,“中国去哪儿?从成员到责任,”美中关系全国委员会。
  10. “非常小”: Adam Segal,“实际参与:为美中贸易划清界限”,《华盛顿季刊》第 27 期,第 3 期(2010 年 1 月 7 日):162。
  11. “经过验证的最终用户”: “中芯国际为美国政府获得经过验证的最终用户身份”,中芯国际,2007 年 10 月 19 日, https://www.smics.com/en/site/news_read/4294
  12. 比竞争对手“跑得更快”:这种共识出现的最好历史是 Hugo Meijer,与敌人交易(牛津大学出版社,2016 年)。
  13. 没有人在听: Van Atta 等人,“全球化和美国半导体工业”,国防分析研究所,2007 年 11 月 20 日,第 2-3 页。

第35章“真正的男人有工厂”

CHAPTER 35 “Real Men Have Fabs”

  1. 最终可能会杀了你: Craig Addison, Silicon Shield(Fusion PR,2001),p。77.
  2. 华丽而成功的推销员: Peter J. Schuyten,“推销员的蜕变”,纽约时报,1979 年 2 月 25 日。
  3. 180 纳米: Varas 等人,“在不确定的时代加强全球半导体供应链”,第 10 页。18.
  4. 需要四分之一的资本投资:同上,p. 17.
  5. 最大的模拟芯片制造商: Peter Clarke,“2020 年十大模拟芯片制造商”, eeNews,2021 年 6 月 3 日。
  6. 这四家公司控制了大约 85%: Joonkyu Kang,“DRAM 产业研究”,硕士论文,麻省理工学院,2010 年,第 10 页。13.
  7. Elpida 挣扎求生: Hiroko Tabuchi,“在日本,PC 芯片制造商破产”,纽约时报,2012 年 2 月 27 日。
  8. 新加坡等国家的政府补贴: Varas 等人,“加强不确定时代的全球半导体供应链”,第 18.
  9. 供应 35% 的市场份额: Ken Koyanagi,“SK-Intel NAND 交易指向更广泛的芯片行业重组”, Nikkei Asia,2020 年 10 月 23 日;“三星电子在平泽增加 NAND 闪存生产线”, Pulse,2020 年 6 月 1 日。
  10. “现在听我说,好好听我说”:约翰·伊斯特,“真正的男人有工厂。Jerry Sanders、TJ Rodgers 和 AMD,” SemiWiki,2019 年 7 月 29 日。

第36章无言革命

CHAPTER 36 The Fabless Revolution

  1. “不是一家真正的半导体公司”: Paul McLellan,“芯片和技术简史”, SemiWiki,2013 年 3 月 19 日, https: //semiwiki.com/eda/2152-a-brief-history-of-芯片和技术/ ; 2021 年采访戈登·坎贝尔。
  2. 英伟达,出身卑微: 2021 年采访 Chris Malachowsky。
  3. 孩提时代: Steve Henn,“科技先锋为硅谷的成功注入了艰辛的教训”,NPR,2012 年 2 月 20 日, https://www.npr.org/sections/alltechconsidered/2012/02/20/147162496/tech-先锋频道 - 硬课程 - 进入硅谷 - 成功
  4. 图形的未来: “Jen-Hsun Huang”,StanfordOnline,YouTube 视频,2011 年 6 月 23 日, https://www.youtube.com/watch?v=Xn1EsFe7snQ
  5. 在这个软件工作上投入了大量资金: Ian Buck,“通用计算 GPU 的演变”,2010 年 9 月 20-23 日, https://www.nvidia.com/content/GTC-2010/pdfs/2275_GTC2010.pdf;Don Clark,“为什么 24 岁的芯片制造商是科技界的热门前景之一”,《纽约时报》,2017 年 9 月 1 日;Pradeep Gupta,“CUDA Refresher:回顾 GPU 计算的起源”,Nvidia,2020 年 4 月 23 日, https://developer.nvidia.com/blog/cuda-refresher-reviewing-the-origins-of-gpu-computing/ .
  6. 发现了一个巨大的并行处理新市场: Ben Thompson,“Apple to Build Own GPU, the Evolution of GPUs, Apple and the General-Purpose GPU”, Stratechery Newsletter,2017 年 4 月 12 日;Ben Thompson,“Nvidia 的集成梦想”,Stratechery 通讯,2020 年 9 月 15 日。
  7. 没有它们就不可能制造手机: Hsiao-Wen Wang,“TSMC Takes on Samsung”, CommonWealth,2013 年 5 月 9 日;Timothy B. Lee,“如何高通在近 20 年内撼动了手机行业”,Ars Technica,2019 年 5 月 30 日。
  8. 数千万行代码: Susie Armstrong 访谈,2021 年。
  9. 尚未制造任何芯片: Daniel Nenni,“高通的详细历史”, SemiWiki,2018 年 3 月 9 日;Joel West,“在高通之前:Linkabit 和圣地亚哥电信业的起源”,圣地亚哥历史杂志 https ://sandiegohistory.org/journal/v55-1/pdf/v55-1west.pdf 。
  10. 专注于他们在频谱管理和半导体设计方面的核心优势: 2021 年对两位高通高管的采访。

第37章张忠谋的大同盟

CHAPTER 37 Morris Chang’s Grand Alliance

  1. 该公司宣布将其芯片设计和制造业务分拆: Michael Kanellos,“时代的终结,AMD 的 Sanders Steps Aside”,CNET,2002 年 4 月 24 日;Peter Bright,“AMD 退出芯片制造业务”, Wired,2012 年 3 月 5 日。
  2. 连台积电都担心: 2021 年蒋尚义专访。
  3. 已经拥有全球大约一半的代工市场: Mark LaPedus,“GlobalFoundries 会成功还是失败?” EE Times,2010 年 9 月 21 日, https://www.eetimes.com/will-globalfoundries-succeed-or-fail/
  4. 一个将自己与大型竞争对手区分开来的机会: Claire Sung 和 Jessie Shen,“台积电 40nm 产量问题重新浮出水面,CEO 承诺在年底前修复”, Digitimes,2009 年 10 月 30 日;Mark LaPedus,“台积电确认 40 纳米良率问题,给出预测”, EE Times,2009 年 4 月 30 日。
  5. 感觉就像一部电梯: 2022 年对 Rick Cassidy 的采访。
  6. 不惜一切代价: Russell Flannery,“Fageless and Peerless in an Era of Fabless”,福布斯,2012 年 12 月 9 日;Hsiao-Wen Wang,“台积电挑战三星”, CommonWealth,2013 年 5 月 9 日。
  7. 围绕台积电合并:王,“台积电挑战三星”。
  8. “有停滞”:弗兰纳里,“无晶圆时代的永恒和绝世”。
  9. 重新获得直接控制权: Lisa Wang,“TSMC Reshuffle Stuns Analysts”,台北时报,2009 年 6 月 12 日;Yin-chuen Wu 和 Jimmy Hsiung,“我愿意从头开始”, CommonWealth,2009 年 6 月 18 日。
  10. “产能过多”: Robin Kwong,“对于台积电来说,产能过剩总比过少好”,《金融时报》,2010 年 6 月 24 日。
  11. “我们才刚刚起步”:弗兰纳里,“无晶圆时代的永恒和无与伦比”。

第38章苹果硅

CHAPTER 38 Apple Silicon

  1. “什么是软件?”: Dag Spicer,“史蒂夫·乔布斯:从车库到世界上最有价值的公司”,计算机历史博物馆,2011 年 12 月 2 日;我是由史蒂夫·切尼(Steve Cheney)指导的,“1980 年:史蒂夫·乔布斯(Steve Jobs)谈硬件和软件的融合”,史蒂夫·切尼(Steve Cheney)—技术、业务和战略,2013 年 8 月 18 日。
  2. 这款革命性的新手机还有许多其他芯片:有关这些 iPhone 1 拆解的详细信息,请参阅 Jonathan Zdziarski,“第 2 章。理解iPhone,”O'Reilly,https://www.oreilly.com/library/view/iphone-forensics/9780596153588/ch02.html;“第一代 iPhone 拆解” , IFIXIT,2007 年 6 月 29 日。
  3. 新 iPad 和 iPhone 4: Bryan Gardiner,“Apple 购买 PA Semi 的四个原因”, Wired,2000 年 4 月 23 日;Brad Stone、Adam Satariano 和 Gwen Ackerman,“你从未听说过的最重要的 Apple 高管”,彭博社,2016 年 2 月 18 日。
  4. 为什么 Apple 的产品运行如此顺利: Ben Thompson,“Apple 不断变化的差异化”, Stratechery,2020 年 11 月 11 日;Andrei Frumusanu,“Apple 宣布 Apple Silicon M1:抛弃 x86——基于 A14 的预期”, AnandTech,2020 年 11 月 10 日。
  5. 全球利润的 60%: Harald Bauer、Felix Grawert 和 Sebastian Schink,“无线通信半导体:行业增长引擎”,麦肯锡公司(2012 年秋季):图表 2。
  6. 中国流水线工人: Harrison Jacobs,“‘iPhone 城’内部,全球一半 iPhone 生产的中国工厂重镇”,《商业内幕》 ,2018 年 5 月 7 日。
  7. 越南和印度也是如此: Yu Nakamura,“富士康准备在印度制造 iPhone 12,从中国转移”,日经亚洲,2021 年 3 月 11 日。

第 39 章EUV

CHAPTER 39 EUV

  1. 2012 年向 ASML 投资 40 亿美元: Dylan McGrath,“英特尔再次削减对 ASML 的股份”, EE Times,2018 年 10 月 12 日。
  2. “解决一个不可能的问题”:约翰·泰勒访谈,2021 年。
  3. 从激光系统中吸热: 2021 年对两名通快高管的采访。
  4. 每秒数百万次: “通快激光放大器”,通快, https ://www.trumpf.com/en_US/products/laser/euv-drive-laser/ 。
  5. 457,329 个零部件:采访两名通快高管,2021 年;Mark Lourie,“II-VI Incorporated 扩大了用于 EUV 光刻中通快高功率 CO2 激光器的金刚石窗口的制造能力”,GlobeNewswire,2018 年 12 月 19 日, https: //www.globenewswire.com/news-release/2018/12/ 19/1669962/11543/en/II-VI-Incorporated-Expands-Manufacturing-Capacity-of-Diamond-Windows-for-TRUMPF-High-Power-CO2-Lasers-in-EUV-Lithography.html
  6. 劳伦斯利弗莫尔国家实验室的研究人员: C. Montcalm,“用于极紫外光刻的多层反射涂层”,能源部科技信息办公室,1998 年 3 月 10 日, https://www.osti.gov/servlets/purl /310916
  7. 打一个远到月球的高尔夫球: “Peter Kurz 博士访谈:'在月球上打高尔夫球'”,“光子学世界” https://world-of-photonics.com/en/newsroom/ photonics-industry-portal/photonics-interview/dr-peter-kuerz/;“蔡司——为明天的微芯片开辟新天地”,蔡司集团,YouTube 视频,2019 年 8 月 2 日, https://www.youtube.com/watch?v=XeDCrlxBtTw
  8. “就像一台机器”: “负责任的供应链:为高科技行业设定更高的标准”,ASML, https ://www.asml.com/en/company/sustainability/responsible-supply-chain ;2021 年采访 Frits van Houts。
  9. 它向蔡司支付了 10 亿美元: “新闻稿:蔡司和 ASML 加强合作,在 2020 年初推出下一代 EUV 光刻技术”,ASML,2016 年 11 月 3 日, https://www.asml.com/en/news/press-发布/2016/zeiss-and-asml-strengthen-partnership-for-next-generation-of-euv-lithography
  10. “如果你不守规矩”: 2021 年采访 ASML 供应商高管。
  11. 三万小时: Igor Fomenkov 等人,“Light Sources for High-Volume Manufacturing EUV Lithography: Technology, Performance, and Power Scaling”, Advanced Optical Technologies 6,第 3-4 期(2017 年 6 月 8 日)。
  12. 打印“x”:这种计算光刻技术的描述借鉴了 Jim Keller,“摩尔定律没有死”,加州大学伯克利分校 EECS 活动,YouTube 视频,2019 年 9 月 18 日, https://www.youtube.com/watch?v =oIG9ztQw2Gc
  13. 美国生产设备的关键部件: “Trumpf Consolidates EUV Lithography Supply Chain with Access Laser Deal”,Optics.org,2017 年 10 月 4 日, https: //optics.org/news/8/10/6 。

第40章“没有B计划”

CHAPTER 40 “There Is No Plan B”

  1. “没有 B 计划”: Anthony Yen,“为大批量制造开发 EUV 光刻技术——个人旅程”, IEEE 技术简报 https: //www.ieee.org/ns/periodicals/EDS/EDS-APRIL-2021 -HTML-V2/InnerFiles/LandPage.html
  2. “他们的配偶不抱怨”:采访蒋尚义,2021。
  3. 不遗余力地测试和改进 EUV 工具: Lisa Wang,“TSMC Stalwart Takes SMIC Role”,台北时报,2016 年 12 月 22 日;Jimmy Hsiung,“蒋尚义:集结部队”, CommonWealth,2007 年 12 月 5 日;2021 年采访蒋尚义和甄子丹。
  4. 收购特许半导体: Timothy Prickett Morgan,“AMD 的 GlobalFoundries 消耗特许半导体竞争对手”,注册,2010 年 1 月 14 日。
  5. 倒置的金字塔: 2021 年 IBM 前高管访谈。
  6. 出售他们的芯片部门: 2021 年对两位半导体高管的采访。
  7. 10% 的代工市场: “Apple 推动了 2015 年台积电的整个代工销售增长”, IC Insights,2016 年 4 月 26 日。
  8. 700,000: “三星、台积电在可用晶圆厂产能方面保持领先”, IC Insights,2016 年 1 月 6 日。这个数字以 200 毫米晶圆计算每月晶圆。当时,行业的前沿正在转向 300 毫米晶圆,每个晶圆可以容纳大约两倍的芯片。因此,以 300 毫米晶圆为基础的晶圆月计算量较低。
  9. 从三星获得 14 纳米工艺许可: Peter Bright,“AMD 完成芯片制造业务的退出”, Wired,2012 年 3 月 5 日。
  10. EUV 计划被取消:采访三位前 GlobalFoundries 高管,其中一位专注于 EUV,2021;关于研发支出,见GlobalFoundries 的 IPO 招股说明书,证券交易委员会,2021 年 10 月 4 日,第 81,https://www.sec.gov/Archives/edgar/data/0001709048/000119312521290644/d192411df1.htm。另请参阅 Mark Gilbert,“2019 年第一季度的第四季度招聘前景依然强劲”,SemiWiki,2018 年 11 月 4 日,https: //semiwiki.com/semiconductor-manufacturers/globalfoundries/7749-globalfoundries-pivot-explained/q 。

第 41 章英特尔如何忘记创新

CHAPTER 41 How Intel Forgot Innovation

  1. 更多磨练工艺的机会: Nick Flaherty,“前五名芯片制造商主导全球晶圆产能”, eeNews,2021 年 2 月 11 日。
  2. 可以达到数百万美元: Or Sharir、Barak Peleg 和 Yoav Shoham,“训练 NLP 模型的成本:简明概述”, AI21 实验室,2020 年 4 月。
  3. 美国最有价值的半导体公司: Wallace Witkowski,“按市值计算,英伟达超越英特尔成为美国最大的芯片制造商”, MarketWatch,2020 年 7 月 8 日。
  4. 更强大的 TPU 的价格: “Cloud TPU 定价”,谷歌云, https://cloud.google.com/tpu/pricing;截至 2021 年 11 月 5 日的价格。
  5. “我基本上一直在经营我们的代工业务”: Chris Nuttall,“Chip Off the Old Block Takes Helm at Intel”,《金融时报》,2013 年 5 月 2 日。
  6. 内部支持很少: 2021 年采访前英特尔代工高管。
  7. 几年后关闭: Dylan McGrath,“英特尔确认为第二次 FPGA 启动的代工厂”, EE Times,2012 年 2 月 21 日。
  8. 几乎没有解释发生了什么问题: Joel Hruska,“英特尔承认其 10nm 计划‘过于激进’”, Extreme Tech,2019 年 7 月 18 日。
  9. 延迟采用 EUV 工具:采访 Pat Gelsinger,彭博社,2021 年 1 月 19 日, https://www.bloomberg.com/news/videos/2022-01-19/intel-ceo-gelsinger-on-year-ahead-全球商业视频
  10. 安装在台积电: Ian Cutress,“台积电:我们有 50% 的 EUV 安装,60% 的晶圆产能”, AnandTech,2020 年 8 月 27 日。

第42章中国制造

CHAPTER 42 Made in China

  1. “没有信息化就没有现代化”: Rogier Creemers 主编,“中央网络安全和信息化领导小组成立”,中国版权与传媒,2014 年 3 月 1 日, https://chinacopyrightandmedia.wordpress.com/2014/ 03/01/central-leading-group-for-internet-security-and-informatization-established/
  2. “进行真正的政治改革”: Evan Osnos,“习的美国之旅”,《纽约客》 ,2012 年 2 月 15 日。
  3. 成千上万的审查员: Katie Hunt 和 CY Xu,“中国雇佣 200 万人来监管互联网”,CNN,2013 年 10 月 7 日。
  4. 《地球村》:罗吉尔·克里默斯主编,习近平,《在网络安全和信息化工作会议上的讲话》,中国版权与传媒,2016年4月19日, https://chinacopyrightandmedia.wordpress.com/2016/04/ 19/speech-at-the-work-conference-for-cybersecurity-and-informatization/翻译调整。
  5. “供应链的‘生命之门’”:同上。
  6. “与英特尔芯片配对”:同上。
  7. 大多数中国电脑:几乎所有PC中的CPU芯片都是由美国的英特尔或AMD设计的,尽管这两家公司都在其他国家生产芯片。
  8. 进口半导体的钱比石油还多:参见联合国 Comtrade 集成电路 (8542) 和石油 (2709) 的数据。
  9. 监控技术: Drew Harwell 和 Eva Dou,“报告称,华为测试了可以识别维吾尔少数民族和警报警察的 AI 软件”,华盛顿邮报,2020 年 12 月 8 日。
  10. 依靠美国公司的芯片: Paul Mozur 和 Don Clark,“中国的监控国家吸收数据。美国科技是对其进行分类的关键,”《纽约时报》,2020 年 11 月 22 日。
  11. KT Li: Morris Chang 的口述历史,计算机历史博物馆。

第43章“发起攻击”

CHAPTER 43 “Call Forth the Assault”

  1. “没有人会成为贸易战的赢家”:安娜·布鲁斯-洛克哈特,“中国国家主席习近平在 2017 年达沃斯的热门语录”,世界经济论坛,2017 年 1 月 17 日, https://www.weforum.org /agenda/2017/01/chinas-xi-jinping-at-davos-2017-top-quotes/
  2. “保护将带来巨大的繁荣和力量”: “全文:2017 年唐纳德特朗普就职演讲稿,” Politico,2017 年 1 月 20 日。
  3. “Xi sounding rather more presidential”: Ian Bremmer, “Xi sounding rather more presidential than US president-elect. #Davos,”推特,2017 年 1 月 17 日, https: //twitter.com/ianbremmer/status/821304485226119169 。
  4. “对全球化的有力防御”: Jamil Anderlini、Wang Feng 和 Tom Mitchell,“习近平在达沃斯为全球化提供强有力的防御”,《金融时报》,2017 年 1 月 17 日;习近平,《习近平在世界经济论坛上的主旨演讲全文》,CGTN,2017年1月17日, https://america.cgtn.com/2017/01/17/full-text-of-xi-jinping-世界经济论坛的主题演讲
  5. “全球化的希望”: Max Ehrenfreund,“世界领导人在民粹主义反抗中在达沃斯找到全球化的希望”,华盛顿邮报,2017 年 1 月 17 日。
  6. “国际社会期待中国”: Isaac Stone Fish,“达沃斯的共产党员”,大西洋,2017 年 1 月 18 日。
  7. 《闯关》: http: //politics.people.com.cn/n1/2016/0420/c1001-28291806.html;克里默斯主编,习近平,“在网络安全和信息化工作会议上的讲话”。
  8. 喜欢现状的政府官员:关于习近平的无能与现状,请参阅 Daniel H. Rosen,“中国的经济清算”,外交事务,2021 年 7 月至 8 月。
  9. “投资规模快速上升”:中国国务院报告,《促进全国集成电路产业发展纲要》, http://www.csia.net.cn/Article/ShowInfo.asp?InfoID=88343 .
  10. 数据由乔治城大学的学者汇总: Saif M. Khan、Alexander Mann 和 Dahlia Peterson,“半导体供应链:评估国家竞争力”,安全和新兴技术中心,2021 年 1 月,第 10 页。8、 https://cset.georgetown.edu/wp-content/uploads/The-Semiconductor-Supply-Chain-Issue-Brief.pdf
  11. 让中国依赖 Nvidia 和 AMD: Saif M. Khan 和 Alexander Mann,“AI 芯片:它们是什么以及它们为何重要”,安全和新兴技术中心,2020 年 4 月,第 29-31 页, https://cset .georgetown.edu/publication/ai-chips-what-they-are-and-why-they-matter/
  12. 到 2025 年将达到 30%: “中国预测将远远低于其‘中国制造 2025’的 IC 目标”, IC Insights,2021 年 1 月 6 日, https://www.icinsights.com/news/bulletins/China-Forecast -To-Far-Far-Short-Of-Its-Made-In-China-2025-Goals-For-ICs/ .
  13. 私营部门投资者流离失所: “博士。周子学就任中芯国际董事长”,新闻稿,中芯国际,2015 年 3 月 6 日, http://www.smics.com/en/site/news_read/4539;Doug Fuller, Paper Tigers,Hidden Dragons(牛津大学出版社,2016 年)描绘了政府影响力增加的早期阶段。
  14. 遍布全国的小型设施: 2021 年采访一家中国代工厂的前 CEO;富勒、纸老虎、潜龙
  15. “‘让我们赔钱’”: 2020 年欧洲半导体高管访谈。
  16. 基金的主要“投资者”:巴里·诺顿,《中国产业政策的崛起,1978 年至 2020 年》(拉丁美洲和加勒比中国学术网络,2021 年),第 10 页。114.
  17. 一种新的“风险投资”模式: Arthur Kroeber,“风险投资国家”, GaveKal Dragonomics,2021 年 3 月。
  18. 只有政府才能赌上这样的赌注: Dieter Ernest,从追赶到锐意进取:中国的半导体政策(东西方中心,2015 年),第 19 页。
  19. 减少外国芯片的份额: Luffy Liu,“倒计时:中国离 40% 的芯片自给率有多近?” EE 时报,2019 年 4 月 11 日。
  20. “红色供应链”: https ://www.cw.com.tw/article/5053334 ;https://www.twse.com.tw/ch/products/publication/download/0003000156.pdf。感谢 Wei-Ting Chen 帮助翻译这些文件。

第 44 章技术转让

CHAPTER 44 Technology Transfer

  1. “一个很好的机会”: David Wolf,“当中国免费获得 IP 时,为什么还要购买硬件?” 外交政策,2015 年 4 月 24 日。
  2. 与美国网络侦探合作: IBM 否认向国家安全局提供任何客户数据;克莱尔·凯恩·米勒(Claire Cain Miller),“美国国家安全局间谍活动让美国科技公司付出代价的启示”,《纽约时报》,2014 年 3 月 21 日;Sam Gustin,“IBM:我们尚未向 NSA 提供任何客户数据”,《时代》 ,2014 年 3 月 14 日。
  3. “一系列非常重要的经济改革”:马修·米勒,“IBM 首席执行官访问中国与政府领导人建立信任会谈:资料来源”,路透社,2014 年 2 月 12 日。
  4. 与中国高级官员会面:参见 2014 年 7 月与北京市长的会晤,IBM 新闻,“今天,#IBM 首席执行官 Ginni Rometty 在#China 的北京会议中心会见了北京市长王安顺。[照片],”推特,7 月 9 日, 2014, https://mobile.twitter.com/ibmnews/status/486873143911669760;2016 年与李克强会面,“IBM 的 Ginni Rometty 会见中国总理李克强”,福布斯,2016 年 10 月 22 日。
  5. 路透社报道:米勒,“IBM 首席执行官访问中国与政府领导人建立信任会谈:消息来源。”
  6. “加强集成电路合作”: “中国副总理会见IBM总裁”,人民网,2014年11月13日, http ://en.people.cn/n/2014/1113/c90883-8808371.html 。
  7. 很少的服务器市场份额: Timothy Prickett Morgan,“X86 服务器主导数据中心——目前”, Next Platform,2015 年 6 月 4 日。
  8. “创造一个充满活力的新生态系统”: Paul Mozur,“IBM Venture with China 引发关注”,纽约时报,2015 年 4 月 19 日。
  9. “巨大的安全风险”:同上。
  10. Qualcomm 收入的一个关键来源: “China Deal Squeezes Royalty Cuts from Qualcomm”, EE Times,2015 年 2 月 10 日。
  11. 华芯通没有业绩记录:陈庆庆,“高通失败的合资企业在竞争加剧的情况下暴露出糟糕的芯片组策略:知情人”,《环球时报》,2019 年 4 月 22 日;Aaron Tilley、Wayne Ma 和 Juro Osawa,“Qualcomm 的中国风险投资显示​​北京科技野心的风险”,信息,2019 年 4 月 3 日;李涛,“高通表示将在中国贵州省农村地区终止与当地政府的芯片合作伙伴关系”,《南华早报》,2019 年 4 月 19 日。
  12. 包括 Phytium: “服务器和云领导者合作创建基于中国的绿色计算联盟”, Arm,2016 年 4 月 15 日, https ://www.arm.com/company/news/2016/04/server-and-cloud-领导者合作创建基于中国的绿色计算联盟
  13. 为 Phytium 工作:参见“Wei Li”,LinkedIn, https ://www.linkedin.com/in/wei-li-8b0490b/?originalSubdomain=cn ;Ellen Nakashima 和 Gerry Shih,“中国利用美国芯片技术构建先进武器系统”,华盛顿邮报,2021 年 4 月 9 日。
  14. “世界级”: “AMD 和南通富士通微电子有限公司关闭半导体组装和测试合资企业”,AMD,2016 年 4 月 29 日,
  15. 中国公司和政府机构:与 AMD 合资的投资者之一是中国科学院,它是中国国家的一部分;请参阅 Ian Cutress 和 Wendell Wilson,“测试中国 x86 CPU:深入了解基于 Zen 的 Hygon Dhyana 处理器”, AnandTech,2020 年 2 月 27 日。
  16. “对微处理器、半导体或中国一无所知”: 2021 年芯片业内人士访谈。
  17. 最终不依赖于金钱: 2021 年采访 Stacy Rasgon。
  18. 简单地调整 AMD 设计:采访一位业内人士和一位前美国官员,2021 年;Don Clark,“AMD 将芯片技术授权给 China Chip Venture”,《华尔街日报》,2016 年 4 月 21 日;Usman Pirzada,“不,AMD 没有出售 x86 王国的密钥——这就是中国联合的方式Venture Works,” Wccftech,2019 年 6 月 29 日;Cutress 和 Wilson,“测试中国 x86 CPU”;Stewart Randall,“AMD 真的放弃了‘王国之钥’吗?” 技术节点,2019 年 7 月 10 日。
  19. 华尔街日报》报道: Kate O'Keeffe 和 Brian Spegele,“美国大型芯片制造商如何为中国提供‘通往王国的钥匙’”,《华尔街日报》,2019 年 6 月 27 日。
  20. 在华盛顿引起人们的注意: “AMD EPYC Momentum 随着腾讯和京东的数据中心承诺、曙光和联想的新产品细节而增长”,新闻稿,AMD,2017 年 8 月 23 日, https: //ir.amd.com/news -events/press-releases/detail/788/amd-epyc-momentum-grows-with-datacenter-commitments-from;采访前美国官员,2021 年。
  21. “核武器和高超音速武器”: Craig Timberg 和 Ellen Nakashima,“超级计算是美中高科技战的最新前沿”,华盛顿邮报,2019 年 6 月 21 日;工业和安全局,“将实体添加到实体列表和修订实体列表条目”,联邦公报,2019 年 6 月 24 日, https://www.federalregister.gov/documents/2019/06/24/ 2019-13245/addition-of-entities-to-the-entity-list-and-revision-of-an-entry-on-the-entity-list;Michael Kan,“美国试图通过阻止出口来阻止中国在 Exascale 超级计算机上的工作”, PC Mag,2021 年 4 月 8 日。
  22. 在“技术、贸易和军民融合听证会:中国对人工智能、新材料和新能源的追求”中的“Elsa Kania 声明”中宣传其与中国军方的联系,《美中经济与安全评论》委员会,2019 年 6 月 7 日,第 69, https://www.uscc.gov/sites/default/files/2019-10/June%207,%202019%20Hearing%20Transcript.pdf
  23. 不确定曙光是如何获得这些芯片的: Anton Shilov,“中国服务器制造商曙光拥有自己的 Radeon Instinct MI50 计算卡(更新)”, Tom's Hardware,2020 年 10 月 15 日, https: //www.tomshardware.com/news /chinese-server-maker-sugon-has-its-own-radeon-instinct-mi50-compute-cards。AMD 代表没有回应我关于其与曙光关系的信息的请求。
  24. 容易受到来自北京的政治压力: Alexandra Alper 和 Greg Roumeliotis,“独家:美国清算软银对 GM 支持的 Cruise 的 22.5 亿美元投资”,路透社,2019 年 7 月 5 日;Dan Primack,“SoftBank 的 CFIUS 解决方法”, Axios,2018 年 11 月 29 日;Heather Somerville,“软银选择与美国国家安全委员会的战斗”,路透社,2019 年 4 月 11 日。
  25. 仅 7.75 亿美元: Cheng Ting-Fang、Lauly Li 和 Michelle Chan,“软银出售 Arm 中国如何播下不和谐的种子”,日经亚洲,2020 年 6 月 16 日;《Arm 中国之战》,《金融时报》,2020 年 6 月 26 日。
  26. 一位 Arm 高管告诉Nikkei Asia Cheng Ting-Fang 和 Debby Wu,“ARM 在中国的合资企业帮助培育‘安全’芯片技术”, Nikkei Asia,2017 年 5 月 30 日。

第45章“合并势在必行”

CHAPTER 45 “Mergers Are Bound to Happen”

  1. 被誉为芯片亿万富翁:广冈信孝,“了解紫光集团,中国芯片战略的关键参与者”,日经亚洲,2020 年 11 月 12 日;“大学的大手笔交易让赵成为筹码亿万富翁”,《中国日报》 ,2015 年 3 月 25 日。
  2. 通往十亿美元财富之路:广冈,《走进清华紫光》;王悦,“认识清华的赵卫国,引领中国芯片雄心的人”,福布斯,2015 年 7 月 29 日。
  3. “私人朋友”: Kenji Kawase,“清华紫光的债券违约是意外吗?” 日经亚洲,2020 年 12 月 4 日;Eva Dou,“中国最大的芯片制造商与惠普价值部门的合作可能高达 50 亿美元”,华尔街日报,2015 年 4 月 15 日;王,《遇见清华的赵卫国》;王悦,“清华引领中国芯片驱动”,日经亚洲,2015 年 7 月 29 日。
  4. 习近平大学室友: Dieter Ernst,“中国半导体的大胆战略——零和博弈还是合作的催化剂?” 东西方中心,2016 年 9 月;Willy Wo-Lap Lam,“习近平集团成员揭晓”,詹姆斯敦基金会,2014 年 2 月 7 日;2008年底,陈曦卸任清华大学校长一职。
  5. “我们所有的交易都是以市场为导向的”:王,“认识清华的赵卫国。”
  6. “也许你会抓到一只鹿”: Dou,“中国最大的芯片制造商可能与惠普价值部门达成高达 50 亿美元的合作。”
  7. 赵花的金额: Zijing Wu 和 Jonathan Browning,“China University Deal Spree Exposes Zhao as Chip Billionaire”,彭博社,2015 年 3 月 23 日。
  8. “巨大的协同效应”: Saabira Chaudhuri,“展讯通信同意以 17.8 亿美元收购”,华尔街日报,2013 年 7 月 12 日。
  9. 赵与英特尔达成交易: “英特尔和紫光集团合作加速基于英特尔的移动设备的开发和采用”,新闻稿,英特尔新闻室,2014 年 9 月 25 日, https://newsroom.intel.com/news-releases /intel-and-tsinghua-unigroup-collaborate-to-accelerate-development-and-adoption-of-intel-based-mobile-devices/#gs.7y1hjm
  10. “国家优先”: Eva Dou 和 Wayne Ma,“英特尔向中国芯片制造商投资 15 亿美元,”华尔街日报,2014 年 9 月 26 日;Cheng Ting-Fang,“英特尔与北京支持的芯片制造商的 5G 调制解调器联盟结束”,日经亚洲,2019 年 2 月 26 日。
  11. 拿 240 亿美元代替: Paul McLellan,“中国的记忆:XMC”, Cadence,2016 年 4 月 15 日, https ://community.cadence.com/cadence_blogs_8/b/breakfast-bytes/posts/china-memory-2 ;“中国清华紫光集团将斥资 300 亿美元建设南京芯片厂”,路透社,2017 年 1 月 19 日;Eva Dou,“Tsinghua Unigroup Acquis Control of XMC in Chinese-Chip Deal”,华尔街日报,2016 年 7 月 26 日。
  12. 房地产和在线赌博:乔什·霍维茨(Josh Horwitz),“分析:中国潜在的芯片宠儿清华紫光集团深受债务和不良赌注的困扰”,路透社,2021 年 1 月 19 日。
  13. 宣布投资计划:窦,“中国最大的芯片制造商可能与惠普价值部门结盟高达 50 亿美元。”
  14. 聘请了几位领先的台湾半导体高管: Josephine Lien 和 Jessie Shen,“联电前 CEO 加入紫光集团”, Digitimes Asia,2017 年 1 月 10 日;Matthew Fulco,“台湾芯片制造商关注中国市场”,台湾商业主题,2017 年 2 月 8 日, https://topics.amcham.com.tw/2017/02/taiwan-chipmakers-eye-china-market/
  15. 追求股权和合资: Debby Wu 和 Cheng Ting-Fang,“Tsinghua Unigroup-SPIL Deal Axed on Policy Worries”, Nikkei Asia,2016 年 4 月 28 日。
  16. 购买岛上的皇冠上的明珠: Peter Clarke,“中国清华对联发科感兴趣”, EE News,2015 年 11 月 3 日。
  17. 建议中国应禁止进口台湾芯片: Simon Mundy,“Taiwan's Chipmakers Push for China Thaw”,《金融时报》,2015 年 12 月 6 日;邹驰,TNL传媒集团,2015年11月3日, https: //www.thenewslens.com/article/30138 。
  18. “如果价格合适”: Cheng Ting-Fang,“如果价格合适,芯片制造商将向中国出售股份”,日经亚洲,2015 年 11 月 7 日。
  19. “保护知识产权不会那么容易”: JR Wu,“中国投资者不应在台湾芯片公司获得董事会席位——台积电总裁”,路透社,2016 年 6 月 7 日。
  20. “携手提升地位”: JR Wu,“台湾联发科表示愿意在芯片领域与中国合作”,路透社,2015 年 11 月 2 日。
  21. “你无法逃避这个问题”: Ben Bland 和 Simon Mundy,“台湾考虑解除中国半导体禁令”,《金融时报》,2015 年 11 月 22 日。
  22. 提出购买美光的想法: Eva Dou 和 Don Clark,“中国国有芯片制造商清华紫光集团以 230 亿美元收购美光”,华尔街日报,2015 年 7 月 14 日。
  23. 美国政府的安全担忧: 2021 年对两名前高级官员的采访。
  24. 延长 37 亿美元的报价: Eva Dou 和 Don Clark,“Arm of China-Controlled Tsinghua to Buy 15% Stake in Western Digital”,华尔街日报,2015 年 9 月 30 日。
  25. “这纯粹是一项金融投资”: Eva Dou 和 Robert McMillan,“中国清华紫光集团收购美国芯片制造商 Lattice 的少量股份”,华尔街日报,2016 年 4 月 14 日。
  26. 投资公布几周后: Ed Lin,“China Inc. Retreats from Lattice Semiconductor”,巴伦周刊,2016 年 10 月 7 日。
  27. 由中国政府谨慎资助: Liana Baker、Koh Gui Qing 和 Julie Zhu,“China Government Money Backs Buyout Firm's Deal for US Chipmaker”,路透社,2016 年 11 月 28 日。中国改革控股,中国政府所有的投资基金,是峡谷桥的主要投资者;见吉田纯子,“中国有想象力吗?EE 时报,2020 年 4 月 14 日。
  28. 同时购买了 Imagination: Nick Fletcher,“Imagination Technologies Jumps 13% as Chinese Firm Takes 3% Stake”, Guardian,2016 年 5 月 9 日。
  29. 所以华盛顿也没有阻止它: “Canyon Bridge Confident Imagination Deal Satisfies UK Government”,《金融时报》,2017 年 9 月 25 日;Turner 等人,“Canyon Bridge Is Said to Ready Imagination Bid Minus US Unit”,彭博社,2017 年 9 月 7 日。
  30. 重组董事会: Nic Fides,“中国采取行动控制 Imagination Technologies Stalls”,《金融时报》,2020 年 4 月 7 日。
  31. 内幕交易: “USA v. Chow”, https://www.corporatedefensedisputes.com/wp-content/uploads/sites/19/2021/04/United-States-v.-Chow-2d-Cir.-Apr. -6-2021.pdf;“美利坚合众国诉 Benjamin Chow,” https://www.justice.gov/usao-sdny/press-release/file/1007536/download;詹妮弗·贝内特(Jennifer Bennett),“峡谷桥创始人的内幕交易信念得到维持”,彭博法律,2021 年 4 月 6 日。
  32. 《必然会发生》:王,《见见清华的赵卫国》。
  33. 获得新的“投资”:张思佳,“中国清华紫光集团签署高达 1500 亿元人民币的融资协议”,路透社,2017 年 3 月 28 日。

第46章华为的崛起

CHAPTER 46 The Rise of Huawei

  1. 华为与中国政府之间的关系:主席 Mike Rogers 和高级会员 CA Dutch Ruppersberger,“关于中国电信公司华为和中兴通讯提出的美国国家安全问题的调查报告”,美国众议院情报常设特别委员会,10 月 8 日, 2012, https://republicans-intelligence.house.gov/sites/intelligence.house.gov/files/documents/huawei-zte%20investigative%20report%20(final).pdf,第 11-25 页。
  2. 合成纤维: William Kirby 等人,“华为:数字冷战交火中的全球科技巨头”,哈佛商学院案例 N-1-320-089,p。2.
  3. 楼宇交换设备: Kirby 等人,“华为”;Jeff Black、Allen Wan 和 Zhu Lin,“习近平的科技仙境遭遇逆风”,彭博社,2020 年 9 月 29 日。
  4. 直接复制: Scott Thurm,“华为承认在路由器软件中复制思科的代码”,华尔街日报,2003 年 3 月 24 日。
  5. 该国的间谍机构认为:汤姆布莱克威尔,“独家:华为打倒了北电吗?企业间谍、盗窃和两大电信巨头的兴衰并行,”《国家邮报》 ,2020 年 2 月 20 日。
  6. 150 亿美元的年度研发预算: Nathaniel Ahrens,“中国的竞争力”,战略与国际研究中心,2013 年 2 月, https://csis-website-prod.s3.amazonaws.com/s3fs-public/legacy_files/files/publication/ 130215_competitiveness_Huawei_casestudy_Web.pdf
  7. 游美:田涛、吴春波,《华为的故事》(Sage Publications Pvt. Ltd.,2016 年),第 10 页。53.
  8. “他们觉得自己落后了一百年”: 2021 年采访前 IBM 顾问和后来的华为员工。
  9. “牺牲是军人的最高事业”:钟瑞峰,“华为的‘狼文化’助其成长,却陷入困境”,纽约时报,2018 年 12 月 18 日。
  10. 学习了斯大林格勒: “斯坦福工程英雄讲座:Morris Chang 与总裁 John L. Hennessy 的对话”,斯坦福在线,YouTube 视频,2014 年 4 月 25 日, https://www.youtube.com/watch?v=wEh3ZgbvBrE
  11. 750 亿美元: Chuin-Wei Yap,“国家支持助力华为全球崛起”,华尔街日报,2019 年 12 月 25 日。
  12. 中国高级官员:阿伦斯,“中国的竞争力”。
  13. “民主党还是共和党”:陶和春波,《华为的故事》,第 10 页。58; Mike Rogers 和 Dutch Ruppersberger,“中国电信公司华为和中兴通讯对美国国家安全问题的调查报告”,美国众议院,2012 年 10 月 8 日, https: //stacks.stanford.edu/file/druid:rm226yb7473 /Huawei-ZTE%20Investigative%20Report%20(FINAL).pdf .
  14. 尽可能多的内部设计: 2021 年采访前 IBM 顾问和华为员工。
  15. 台积电的第二大客户:程廷芳和 Lauly Li,“台积电在美国收紧限制后暂停新的华为订单”,日经亚洲,2020 年 5 月 18 日。

第 47 章5G 的未来

CHAPTER 47 The 5G Future

  1. 切换衣柜大小的装备: 2021 年采访肯·汉克勒。
  2. 使用更少功率的同时更精确: 2021 年采访戴夫·罗伯逊。
  3. “像智能手机”: Spencer Chin,“拆解揭示特斯拉 S 像智能手机”, Power Electronics,2014 年 10 月 28 日。
  4. 高质量和具有竞争力的价格: Ray Le Maistre,“BT 的 McRae:华为是‘目前唯一真正的 5G 供应商’”, Light Reading,2018 年 11 月 21 日。
  5. 对华为无线电部门的一项研究:松本纪夫和渡边直树,“华为的基站拆解显示对美国制造的零件的依赖”,日经亚洲,2020 年 10 月 12 日。

第48章下一个偏移

CHAPTER 48 The Next Offset

  1. “心理核攻击”:刘真,“中美对抗:海湾战争如何引发北京的军事革命”,《南华早报》,2021 年 1 月 18 日;另见 Harlan W. Jencks,“中国对‘沙漠风暴’的评估:对中国安全的影响”,《东亚事务杂志》,第 6 期,第 2 期(1992 年夏季/秋季):447-477。
  2. “中国可能超过美国”: “最终报告”,国家人工智能安全委员会,第 10 页。25.
  3. “人工智能武器”: Elsa B. Kania,“中国军事创新中的‘人工智能武器’”,全球中国,布鲁金斯学会,2020 年 4 月。
  4. “三合会”: Ben Buchanan,“人工智能三合会及其对国家安全战略的意义”,安全与新兴技术中心,2020 年 8 月。
  5. 中国在收集数据方面没有任何内在优势: Matt Sheehan,“Much Ado About Data: How America and China Stack Up”,MacroPolo,2019 年 7 月 16 日, https: //macropolo.org/ai-data-我们中国/?rp=e
  6. 59% 的世界顶级AI研究人员: “全球 AI 人才追踪器”,MacroPolo, https: //macropolo.org/digital-projects/the-global-ai-talent-tracker/ 。
  7. 95% 的 GPU: “中国计算能力发展指数白皮书”,tr。Jeffrey Ding,中国信息通信研究院,2021年9月, https ://docs.google.com/document/d/1Mq5vpZQe7nrKgkYJA2-yZNV1Eo8swh_w36TUEzFWIWs/edit# ,中文原文出处: http: //www.caict.ac.cn /kxyj/qwfb/bps/202109/t20210918_390058.htm
  8. 乔治城大学的研究人员: Ryan Fedasiuk、Jennifer Melot 和 Ben Murphy,“利用闪电:中国军队如何采用人工智能”,CSET,2021 年 10 月, https ://cset.georgetown.edu/publication/harnessed-lightning/ ,特别是。fn 84; 关于军民融合,见 Elsa B. Kania 和 Lorand Laskai,“中国军民融合战略的神话和现实”,新美国安全中心,2021 年 1 月 28 日。
  9. 决定性的技术优势: Gian Gentile、Michael Shurkin、Alexandra T. Evans、Michelle Grise、Mark Hvizda 和 Rebecca Jensen,“第三次抵消的历史,2014-2018”,兰德公司,2021;“副部长关于第三次抵消战略工作的讲话”,美国国防部,2016 年 4 月 28 日。
  10. “分布在战场上的计算机”: “DARPA 将马赛克战争的愿景拼凑在一起”,国防高级研究计划局, https://www.darpa.mil/work-with-us/darpa-tiles-together-a-vision马赛克战争
  11. “人机协作”: “设计敏捷的人机团队”,国防高级研究计划局,2016 年 11 月 28 日, https://www.darpa.mil/program/2016-11-28
  12. 据报道,俄罗斯政府还阻碍 GPS 信号: Roger N. McDermott,“Russia's Electronic Warfare Capabilities to 2025”,国际国防与安全中心,2017 年 9 月;“研究地图‘广泛的俄罗斯 GPS 欺骗’”,BBC 新闻,2019 年 4 月 2 日。
  13. 替代导航系统: “自适应导航系统 (ANS)(存档)”,国防高级研究计划局, https://www.darpa.mil/program/adaptable-navigation-systems
  14. 军队的观察和沟通能力:布莱恩·克拉克和丹·帕特,“美国需要一个战略来保护微电子——不仅仅是资金,”哈德逊研究所,2021 年 3 月 15 日。
  15. 电子复兴计划: “DARPA 电子复兴计划”,国防高级研究计划局,2021 年 6 月 28 日, https://www.darpa.mil/work-with-us/electronics-resurgence-initiative
  16. 晶体管结构:关于 FinFET,请参阅 Tekla S. Perry,“FinFET 之父如何帮助拯救摩尔定律”, IEEE Spectrum,2020 年 4 月 21 日。
  17. 10-15%: Norman J. Asher 和 Leland D. Strom,“国防部在集成电路发展中的作用”,国防分析研究所,1977 年 5 月,p。74.
  18. 几亿美元: Ed Sperling,“那个芯片要多少钱?” 半导体工程,2014 年 3 月 27 日。
  19. Spectre and Meltdown: Cade Metz 和 Nicole Perlroth,“研究人员发现世界计算机中的两个主要缺陷”,纽约时报,2018 年 1 月 3 日。
  20. 在通知美国政府之前: Robert McMillan 和 Liza Lin,“英特尔在美国政府之前警告中国公司注意芯片缺陷”,华尔街日报,2018 年 1 月 28 日。
  21. “零信任”: Serge Leef,“电子防御 (SHIELD) 的供应链硬件完整性(存档)”,国防高级研究计划局, https ://www.darpa.mil/program/supply-chain-hardware-integrity- for-electronics-defense#:~:text=%20目标%20of%20DARPA的%20SHIELD,消耗%20to%20be%20cost%20有效;“DARPA 可信微电子方法”, https://www.darpa.mil/attachments/ATrustthroughTechnologyApproach_FINAL.PDF
  22. 现在押注未来: “工作副书记关于第三次抵消战略的讲话。”
  23. “和我们一起在车里”: 2021 年对美国前官员的采访;Gian Gentile、Michael Shurkin、Alexandra T. Evans、Michelle Grise、Mark Hvizda 和 Rebecca Jensen,“第三次抵消的历史,2014-2018 年。”

第49章“我们竞争的一切”

CHAPTER 49 “Everything We’re Competing On”

  1. “明显的恐惧感”: 2021 年对美国前高级官员的采访。
  2. “要埋葬我们”:同上。
  3. 没有将筹码视为重要问题:同上。
  4. 普利兹克发表了一个高调的演讲: “美国商务部长彭妮·普利兹克在战略与国际研究中心发表关于半导体的重要政策演讲”,美国商务部彭妮·普利兹克的演讲,2016 年 11 月 2 日。
  5. 发表了一份报告: “确保美国在半导体领域的长期领导地位”,向总统提交的报告,总统科学技术顾问委员会,2017 年 1 月。
  6. 花了几十年的时间与指控作斗争: Mike Rogers 和 Dutch Ruppersberger,“关于中国电信公司华为和中兴通讯提出的美国国家安全问题的调查报告”,美国众议院,2012 年 10 月 8 日;Kenji Kawase,“ZTE 鲜为人知的根源:中国科技公司从 Grace 手中陨落”, Nikkei Asia,2018 年 4 月 27 日;Nick McKenzie 和 Angus Grigg,“法院文件称,中国的中兴通讯是为了间谍和贿赂而建立的”,《悉尼先驱晨报》,2018 年 5 月 31 日;Nick McKenzie 和 Angus Grigg,“中国腐败公司入围 Telstra 候选名单”,《悉尼先驱晨报》,2018 年 5 月 13 日;“中兴通讯在 2006 年国际 CDMA 市场上名列前茅”,CIOL 局,https://web.archive.org/web/20070927230100/http://www.ciol.com/ciol-techportal/Content/Mobility/News/2007/20703081355.asp
  7. 被指控违反: Juro Osawa 和 Eva Dou,“US to Place Trade Restrictions on China's ZTE”,华尔街日报,2016 年 3 月 7 日;Paul Mozur,“美国就其在伊朗和朝鲜的交易传唤华为”,《纽约时报》,2016 年 6 月 2 日。
  8. 选择惩罚公司:采访两名奥巴马政府官员,2021 年;Osawa 和 Dou,“美国将对中国的中兴实施贸易限制”。
  9. 在他们生效之前:工业和安全局,“从实体名单中删除某些人;将人员添加到实体列表;和 EAR 符合性变更”,《联邦公报》,2017 年 3 月 29 日, https://www.federalregister.gov/documents/2017/03/29/2017-06227/removal-of-certain-persons-from-the-entity-将一个人的列表添加到实体列表和耳朵;Brian Heater,“中兴对违反伊朗制裁表示认罪,同意处以 8.92 亿美元的罚款”, TechCrunch,2017 年 3 月 7 日。
  10. 特朗普一再攻击中国“敲诈我们”: Veronica Stracqualursi,“特朗普 10 次攻击中国及其与美国的贸易关系”,ABC 新闻,2017 年 11 月 9 日。
  11. “你无能为力”:对四位前高级官员的采访,2021 年。
  12. “我们在二十一世纪竞争的一切”: 2021 年前高级官员访谈。
  13. 政府开始关注半导体:同上。
  14. Krzanich 面临强烈反对: Lucinda Shen,“Donald Trump 的推文引发英特尔 CEO 退出商业委员会”,《财富》,2017 年 11 月 9 日;Dawn Chmielewski 和 Ina Fried,“英特尔的 CEO 计划,然后废弃,唐纳德特朗普筹款活动”,CNBC,2016 年 6 月 1 日。
  15. “我们的头号竞争对手”: 2021 年对前高级政府官员的采访。
  16. 太多的技术泄露: 2021 年对三位前高级官员的采访。
  17. 主要是贸易问题: Chad Bown、Euijin Jung 和 Zhiyao Lu,“特朗普、中国和关税:从大豆到半导体”, Vox EU,2018 年 6 月 19 日。
  18. 违反了认罪协议的条款: Steve Stecklow、Karen Freifeld 和 Sijia Jiang,“随着紧张局势升级,美国对中国中兴通讯的销售禁令开启了新的战线”,路透社,2018 年 4 月 16 日。
  19. “几乎无人知晓”: 2021 年对高级政府官员的采访。
  20. “在中国失去太多工作”: Dan Strumpf 和 John D. McKinnon,“特朗普将生命线延伸到受制裁的科技公司中兴通讯”,华尔街日报,2018 年 5 月 13 日;斯科特·霍斯利 (Scott Horsley) 和斯科特·纽曼 (Scott Neuman),“特朗普总统暂停‘美国优先’以挽救中国的就业机会”,NPR,2018 年 5 月 14 日。

第50章福建晋华

CHAPTER 50 Fujian Jinhua

  1. “清除计算机数据”:此帐户来自“美国诉联合微电子公司等,被告”,美国加利福尼亚州北区地方法院,2018 年 9 月 27 日, https: //www.justice.gov/opa/press-release/file/1107251/download和“MICRON TECHNOLOGY, INC. 的投诉”。作为与美国政府和解的一部分,UMC 对这些指控表示认罪。涉案联电员工被台湾法院以刑事罪名定罪、罚款、入狱;公共事务办公室,“台湾公司承认在涉及中国国有企业的刑事案件中盗窃商业秘密罪名成立”,美国部司法委员会,2020 年 10 月 28 日,https://www.justice.gov/opa/pr/taiwan-company-pleads-guilty-trade-secret-theft-criminal-case-involving-prc-state-owned
  2. 超过 50 亿美元: Chuin-Wei Yap 和 Yoko Kubota,“美国禁令威胁北京作为科技强国的野心”,华尔街日报,2018 年 10 月 30 日。
  3. 获得约 7 亿美元: Chuin-Wei Yap,“美光禁止在中国销售某些产品”,华尔街日报,2018 年 7 月 4 日。
  4. 不涉足DRAM行业:在福建晋华案的辩护中,联电强调了其先前的存储芯片专业知识,但其2016年年报却强调“我们……不打算进入DRAM行业”。请参阅向美国证券交易委员会提交的 UMC 表格 20-F,2016 年,第 27.
  5. 900 个文件: Paul Mozur,“在美国芯片设计的抢劫中,中国竞标科技强国”,纽约时报,2018 年 6 月 22 日。
  6. 窃听王的手机:同上。
  7. 禁止美光销售: Yap,“美光禁止在中国销售某些产品。”
  8. “欺负落后国家”: https ://www.storm.mg/article/1358975?mode=whole ,tr。陈伟霆。
  9. 迅速重启: David E. Sanger 和 Steven Lee Meyers,“中断后,中国加速网络间谍活动以获取美国技术”,《纽约时报》,2018 年 11 月 29 日。
  10. 仍然保密: Advanced Micro-Fabrication Equipment Inc.,“AMEC 在涉及 Veeco Instruments (Shanghai) Co. Ltd. 的专利侵权纠纷中赢得禁令”,美通社,2017 年 12 月 8 日, https://www.prnewswire.com/news -releases/amec-wins-injunction-in-patent-infringement-dispute-involving-veeco-instruments-shanghai-co-ltd-300569295.html;Mark Cohen,“半导体专利诉讼第 2 部分:民族主义、透明度和法治”,中国知识产权,2018 年 7 月 4 日, https://chinaipr.com/2018/07/04/semiconductor-patent-litigation-part-2 -民族主义-透明度和法治/; “Veeco Instruments Inc.,原告,针对 SGL Carbon, LLC 和 SGL Group SE,被告,”美国纽约东区地方法院,https://chinaipr2.files.wordpress.com/2018/07/uscourts- nyed-1_17-cv-02217-0.pdf
  11. 主张实施金融制裁: Kate O'Keeffe,“美国采取新的作战计划打击中国窃取商业机密”,华尔街日报,2018 年 11 月 12 日。
  12. 特朗普政府相信东京支持采取强硬举措: 2019-2021 年在华盛顿和东京采访五名政府官员。
  13. “我们他妈的为什么不使用这个?”: 2021 年对前高级官员的采访。
  14. 陷入停顿: James Politi、Emily Feng 和 Kathrin Hille,“美国针对中国芯片制造商的安全问题”,《金融时报》,2018 年 10 月 30 日。

第51章对华为的攻击

CHAPTER 51 The Assault on Huawei

  1. “他们无所不知”: Dan Strumpf 和 Katy Stech Ferek,“美国加强对华为获取芯片的限制”,华尔街日报,2020 年 8 月 17 日。
  2. “不让他们收获果实”: Turpin 在 Elizabeth C. Economy 中引用的《中国的世界》(Wiley,2021 年)。
  3. “我们做错一切的代表”: 2021 年对两名特朗普政府高级官员的采访。
  4. 特恩布尔给自己买了一本 474 页的书: Peter Hartcher,《红区:中国的挑战和澳大利亚的未来》(Black Inc.,2021 年),第 18-19 页。
  5. 逮捕了一名前公司高管: Alicja Ptak 和 Justyna Pawlak,“波兰审判开始于与华为有关的中国间谍案”,路透社,2021 年 6 月 1 日。
  6. 悄悄地施加了严格的限制: Mathieu Rosemain 和 Gwenaelle Barzic,“独家:到 2028 年,法国对华为 5G 设备的限制相当于事实上的禁令”,路透社,2020 年 7 月 22 日。
  7. “后果”: Katrin Bennhold 和 Jack Ewing,“在华为之战中,中国威胁德国的“痛处”:汽车制造商,”纽约时报,2020 年 1 月 16 日。
  8. 华为网络安全实践的缺陷: Gordon Corera,“华为‘未能提高英国安全标准’”,BBC 新闻,2020 年 10 月 1 日。
  9. “中国将成为全球科技强国”:罗伯特·汉尼根,“全面禁止华为等中国科技公司毫无意义”,《金融时报》,2019 年 2 月 12 日。
  10. 违反美国对伊朗的制裁: Shayna Jacobs 和 Amanda Coletta,“孟晚舟可以返回中国,承认帮助华为在伊朗隐瞒交易”,《华盛顿邮报》 ,2021 年 9 月 24 日。
  11. “扼杀华为”: James Politi 和 Kiran Stacey,“美国通过更严格的华为控制升级中国的紧张局势”,《金融时报》,2020 年 5 月 15 日。
  12. “武器化的相互依存”: Henry Farrell 和 Abraham L. Newman,“武器化的相互依存:全球经济网络如何塑造国家强制”,《国际安全》 ,第 44 期,第 1 期(2019 年):42-79。
  13. 进一步加强对华为的限制: “Commerce Addresses Huawei's Efforts to Undermine Entity List, Restricts Products Design and Produced with US Technologies”,美国商务部,2020 年 5 月 15 日, https://2017-2021.commerce.gov/news/press -releases/2020/05/commerce-addresses-huaweis-efforts-undermine-entity-list-restricts.html
  14. 还有它的精神: Kathrin Hille 和 Kiran Stacey,“台积电与美国对华为的出口管制保持一致”,《金融时报》,2020 年 6 月 9 日。
  15. 华为被迫剥离: “由于美国列入黑名单,华为表示将出售关键服务器部门”,彭博社,2021 年 11 月 2 日。
  16. 无法获得必要的芯片: Craig S. Smith,“华为之战如何改变 5G 的面貌”, IEEE Spectrum,2021 年 9 月 29 日。
  17. 由于芯片短缺而延迟: Lauly Li 和 Kenji Kawase,“随着美国遏制措施的开始,华为和中兴通讯减缓了中国 5G 的推出”,日经亚洲,2020 年 8 月 19 日。
  18. ASML 向中国公司提供 EUV 机器: Alexandra Alper、Toby Sterling 和 Stephen Nellis,“特朗普政府向荷兰人施压,要求取消中国芯片设备销售:资料来源”,路透社,2020 年 1 月 6 日。
  19. 曙光:工业和安全局,“将实体添加到实体列表和修订实体列表条目”,联邦公报,2019 年 6 月 24 日,https://www.federalregister.gov/documents/2019/06/24/2019-13245/addition-of-entities-to-the-entity-list-and-revision-of-an-entry-on-the-实体列表
  20. Phytium: Ellen Nakashima 和 Gerry Shih,“中国使用美国芯片技术构建先进武器系统”,华盛顿邮报,2021 年 4 月 9 日。
  21. “不可靠实体清单”:钟山,《商务部令2020年第4号关于不可靠实体清单规定的规定》,中华人民共和国商务部令,2020年9月19日, http://english. mofcom.gov.cn/article/policyrelease/questions/202009/20200903002580.shtml
  22. “美好的事物”: 2021 年对美国前高级官员的采访。

第52章中国的人造卫星时刻?

CHAPTER 52 China’s Sputnik Moment?

  1. 由于该国其他地区仍处于冻结状态: Cheng Ting-Fang 和 Lauly Li,“中国的芯片行业如何对抗冠状病毒封锁”,日经亚洲,2020 年 3 月 18 日。
  2. “推动了北京对科技主导地位的追求”: Dan Wang,“中国的人造卫星时刻?” 外交事务,2021 年 7 月 29 日。
  3. “芯片沙皇”: “习近平挑选上将领导中国与美国的芯片大战”,彭博社,2021 年 6 月 16 日。
  4. 花费数十亿美元补贴:不应认真对待有关中国准备花费高达 1.4 万亿美元补贴技术的新闻头条。北京已批准名义价值约1.5万亿美元的产业“引导基金”,主要由地方当局筹集和使用。然而,这些不仅仅专注于技术。官方指导方针允许这些资金不仅用于“战略性新兴产业”,还用于基础设施和社会住房。因此,就像中国的许多投资项目一样,这笔资金的一部分很有可能最终只是用于补贴更多的房地产开发,而不是支持半导体。黄天磊,“中国政府引导基金:国家产业政策的融资工具”, PIIE,2019 年 6 月 17 日,https://www.piie.com/blogs/china-economic-watch/government-guided-funds-china-financing-vehicles-state-industrial-policy;唐子怡和薛晓莉,“关于中国 6700 亿美元政府引导基金的四件事”,财新环球,2020 年 2 月 25 日。
  5. “没有经验,没有技术,没有才华”:邱晓芬和苏建勋的HSMC调查,杨轩,主编,tr。亚历山大·博伊德(Alexander Boyd),乔丹·施耐德(Jordan Schneider),“十亿美元的抢劫:诈骗者如何利用中国的芯片繁荣致富”, ChinaTalk,2021 年 3 月 30 日, https: //chinatalk.substack.com/p/billion-dollar-heist-how-scammers ; 罗国平和莫业林,“武汉陷入困境的 185 亿美元芯片制造项目并不像地方官员声称的那样特殊”,财新环球,2020 年 9 月 4 日。
  6. 每台机器成本 3 亿美元: Toby Sterling,“英特尔在 Quest for Chipmaking Edge 中以超过 3.4 亿美元的价格订购 ASML 系统”,路透社,2022 年 1 月 19 日。
  7. 从美国搬到瑞士: David Manners,“RISC-V 基金会搬到瑞士”,《电子周刊》 ,2019 年 11 月 26 日。
  8. 政府补贴可能有助于它赢得业务: Dylan Patel,“中国建造了世界上最昂贵的碳化硅工厂,但数字没有加起来”, SemiAnalysis,2021 年 9 月 30 日, https: //semianalysis.com/china-has -建造世界上最昂贵的碳化硅工厂,但数字不加/
  9. 估计表明中国在制造中的份额将会增加: Varas 等人,“政府激励措施和美国在半导体制造方面的竞争力”。
  10. “实现中国梦”: Cheng Ting-Fang 和 Lauly Li,“中国芯片行业如何对抗冠状病毒封锁”,日经亚洲,2020 年 3 月 18 日。

第 53 章短缺和供应链

CHAPTER 53 Shortages and Supply Chains

  1. “我们必须加强我们的游戏”: “拜登总统在半导体和供应链弹性虚拟 CEO 峰会上的讲话”,白宫,2021 年 4 月 12 日;Alex Fang 和 Yifan Yu,“美国再次引领世界,拜登在半导体峰会上告诉 CEO”,日经亚洲,2021 年 4 月 13 日。
  2. 2100 亿美元: AAPC 提交给 BIS 商务部半导体供应链审查,2021 年 4 月 5 日;Michael Wayland,“芯片短缺预计将在 2021 年使汽车行业损失 2100 亿美元的收入”,CNBC,2021 年 9 月 23 日。
  3. 增长 13%: “半导体单位预测将在 2021 年再次超过 1 万亿个设备”, IC Insights,2021 年 4 月 7 日, https://www.icinsights.com/news/bulletins/Semiconductor-Units-Forecast-To-Exceed- 1-Trillion-Devices-Again-In-2021/
  4. “与行业、盟友和合作伙伴”: “情况说明书:拜登-哈里斯政府宣布供应链中断工作组”,2021 年 6 月 8 日, https://www.whitehouse.gov/briefing-room/statements-releases/2021 /06/08/fact-sheet-biden-harris-administration-announces-supply-chain-disruptions-task-force-to-address-short-term-supply-chain-discontinuities/
  5. “作为一个团队与企业合作”: Kotaro Hosokawa,“三星将韩国驻军城市转变为芯片制造繁荣之城”,日经亚洲,2021 年 6 月 20 日。
  6. “经济因素”: Jiyoung Sohn,“三星将投资 2050 亿美元用于芯片,生物技术扩张”,华尔街日报,2021 年 8 月 24 日;Song Jung-a 和 Edward White,“韩国总理支持被假释的三星首席执行官李在镕提前重返工作岗位”,《金融时报》,2021 年 8 月 30 日。
  7. 限制 EUV 工具的转让: Stephen Nellis、Joyce Lee 和 Toby Sterling,“独家:美中科技战让 SK 海力士的关键芯片工厂计划蒙上阴影”,路透社,2021 年 11 月 17 日。
  8. 使台湾出口更具竞争力: Brad W. Setser,“Shadow FX Intervention in Taiwan:解决 100+ 亿美元之谜(第 1 部分)”,外交关系委员会,2019 年 10 月 3 日。
  9. 欧盟领导人建议: “专员蒂埃里·布雷顿在汉诺威工业博览会数字日上的讲话”,欧盟委员会,2020 年 7 月 15 日。
  10. 与索尼合作: Cheng Ting-Fang 和 Lauly Li,“台积电表示将与索尼建立第一家日本芯片厂”,日经亚洲,2021 年 11 月 9 日。
  11. “我们可以决定晶圆厂的位置”: Christiaan Hetzner,“英特尔首席执行官表示,计划在欧洲建设的‘大宏金’晶圆厂将成为世界上最先进的工厂”,《财富》,2021 年 9 月 10 日;Leo Kelion,“英特尔首席执行长 Pat Gelsinger:亚洲制造的芯片太多”,BBC 新闻,2021 年 3 月 24 日。

第54章台湾困境

CHAPTER 54 The Taiwan Dilemma

  1. “您的客户是否担心”: “编辑的成绩单:2330.TW - 2021 年第二季度台积电财报电话会议,”,2021 年 7 月 15 日, https: //investor.tsmc.com/english/encrypt/files/encrypt_file /reports/2021-10/44ec4960f6771366a2b992ace4ae47566d7206a6/TSMC%202Q21%20transcript.pdf
  2. 边走边开枪:刘玄尊,《美军飞机登陆台湾岛后解放军举行海滩突击演习》,《环球时报》,2021年7月18日。
  3. “坚决维护国家主权”:刘玄尊,《在美国连续军事挑衅下,解放军在中国所有主要海域举行演习》,《环球时报》,2021年7月20日。
  4. 软香肠: Chris Dougherty、Jennie Matuschak 和 Ripley Hunter,“毒蛙战略”,新美国安全中心,2021 年 10 月 26 日。
  5. 艰难的斗争: “涉及中华人民共和国的军事和安全发展”,向国会提交的年度报告,国防部长办公室,2020 年,第 3 页。114.
  6. 位于中国领土上的系统: Lonnie Henley,“台湾冲突中的解放军作战概念和重心”,在美中经济与安全审查委员会关于海峡两岸威慑的听证会上的证词,2021 年 2 月 18 日。
  7. “非和平手段”: Michael J. Green,“什么是美国的‘一个中国’政策,它为什么重要?” 战略与国际研究中心,2017 年 1 月 13 日。
  8. 无法获得足够的半导体: Debby Wu,“Chip Linchpin ASML 加入汽车制造商警告恶性循环”,彭博社,2022 年 1 月 19 日。
  9. 蔡英文最近辩称:蔡英文,“台湾与民主斗争”,外交事务,2021 年 11 月至 12 月。
  10. 大多数台湾人认为: Sherry Hsiao,“大多数人认为两岸战争不太可能:民意调查”,台北时报,2020 年 10 月 21 日。
  11. 由于采购半导体问题而造成的严重延误: Ivan Cheberko,“Kosmicheskii Mashtab Importozameshcheniia”, Vedomosti,2020 年 9 月 27 日。
  12. 充满外国微电子: Jack Watling 和 Nick Reynolds,“Z 行动:帝国妄想的垂死挣扎”,皇家联合军种研究所,2022 年 4 月 22 日,第 10-12 页。
  13. 多达 95% 的俄罗斯弹药: Michael Simpson 等人,“通往大马士革之路:俄罗斯在叙利亚的空中战役”,兰德公司,RR-A1170-1,2022 年,第 80.
  14. 超过 200 种半导体: Rebecca Shabad,“拜登在参观阿拉巴马武器厂时强调需要继续武装乌克兰”,CNBC,2022 年 5 月 3 日。
  15. 切断克里姆林宫: Sebastian Moss,“英特尔和 AMD 停止向俄罗斯销售芯片,台积电加入制裁”,数据中心动态,2022 年 2 月 28 日, https://www.datacenterdynamics.com/en/news/intel- and-amd-halt-chip-sales-to-russia-tsmc-joins-in-on-sanctions/
  16. 用于洗碗机的芯片: Jeanne Whalen,“制裁迫使俄罗斯在军用装备中使用电器零件”,华盛顿邮报,2022 年 5 月 11 日。
  17. “我们必须抓住台积电”: “如果美国加大制裁力度,顶级经济学家敦促中国抓住台积电”,彭博新闻,2022 年 6 月 7 日。
  18. 2021 年,这些空军基地进行了升级: Keoni Everington,“中国扩大了离台湾最近的 2 个空军基地”,台湾新闻,2021 年 3 月 8 日;Minnie Chan,“面对台湾暗示战争计划的中国军事空军基地升级”,《南华早报》,2021 年 10 月 15 日;“中国离台湾最近的三个空军基地正在进行重大建设”, Drive,2021 年 10 月 13 日。

结论

Conclusion

  1. 由半导体材料制成: Jack Kilby,“集成电路的发明”, IEEE Transactions on Electron Devices 23,第 7 期(1976 年 7 月):650。
  2. 工程师如 Weldon Word: Paul G. Gillespie,“精确制导弹药:制造比原子弹更有效的炸弹”,利哈伊大学博士论文,第 115. 根据他死后可用的 LinkedIn 页面,Word 似乎于 1953 年开始在 TI 工作。我无法确认这一点。
  3. “家用电脑”: Gordon E. Moore,“Cramming More Components on Integrated Circuits”, Electronics 38,第 8 期(1965 年 4 月 19 日)。
  4. 比人体中的细胞还多: Dan Hutcheson,“图形:晶体管生产已达到天文规模”, IEEE Spectrum,2015 年 4 月 2 日。
  5. “我……想要……变得……变得……富有”:迈克尔·马龙,英特尔三位一体(迈克尔·柯林斯,2014 年),第 3 页。31.
  6. 宣布摩尔定律死亡: John Hennessy,“摩尔定律的终结和更快的通用处理器,以及一条新的前进道路”,美国国家科学基金会,CISE 杰出讲座,2019 年 11 月 22 日, https: //www.nsf.gov /events/event_summ.jsp?cntn_id=299531&org=NSF
  7. 十年后的抨击: Andrey Ovsyannikov,“来自英特尔的更新:洞察英特尔在 HPC 和 AI 方面的创新”,英特尔,2019 年 9 月 26 日, https ://www2.cisl.ucar.edu/sites/default/files/Ovsyannikov %20-%20MC9%20-%20Presentation%20Slides.pdf
  8. “激进的想法”: Gordon E. Moore,“没有永远的指数:但‘永远’可以延迟!” IEEE 国际固态电路会议,2003 年。
  9. 晶体管数量的一百倍: Hoeneisen 和 Mead,“微电子学的基本限制”,第 819-829 页;Scotten Jones,“台积电和三星 5nm 比较”, SemiWiki,2019 年 5 月 3 日, https: //semiwiki.com/semiconductor-manufacturers/samsung-foundry/8157-tsmc-and-samsung-5nm-comparison/ 。
  10. 一条通往密度增加 50 倍的明确道路: “Jim Keller:摩尔定律没有死”,加州大学伯克利分校 EECS 活动,YouTube 视频,2019 年 9 月 18 日,22:00, https://www.youtube.com/观看?v=oIG9ztQw2Gc
  11. “慢车道”: Neil C. Thompson 和 Svenja Spanuth,“计算机作为通用技术的衰落:为什么深度学习和摩尔定律的终结正在分割计算”,工作论文,麻省理工学院,2018 年 11 月, https:// ide.mit.edu/wp-content/uploads/2018/11/SSRN-id3287769.pdf
  12. 结合不同类型的芯片: “异构计算:没有人谈论的范式转变”, Fabricated Knowledge,2020 年 2 月 19 日, https://www.fabricatedknowledge.com/p/heterogeneous-compute-the-paradigm
  13. 在雪堆中唱歌: Kevin Xu,“Morris Chang 的最后演讲”,互联,2021 年 9 月 12 日, https://interconnected.blog/morris-changs-last-speech/

指数

Index

关于索引的说明:此索引中引用的页数是指印刷版的页码。单击页码会将您带到与印刷版中该页开头相对应的电子书位置。如需任何单词或短语位置的完整列表,请使用阅读系统的搜索功能。

A note about the index: The pages referenced in this index refer to the page numbers in the print edition. Clicking on a page number will take you to the ebook location that corresponds to the beginning of that page in the print edition. For a comprehensive list of locations of any word or phrase, use your reading system’s search function.

标志:斯克里布纳

斯克里布纳

Scribner

Simon & Schuster, Inc. 的印记

An Imprint of Simon & Schuster, Inc.

美洲大道 1230 号

1230 Avenue of the Americas

纽约,纽约 10020

New York, NY 10020

www.SimonandSchuster.com

www.SimonandSchuster.com

版权所有 © 2022 克里斯托弗·米勒

Copyright © 2022 by Christopher Miller

保留所有权利,包括以任何形式复制本书或其部分的权利。如需信息,请联系 Scribner Subsidiary Rights Department, 1230 Avenue of the Americas, New York, NY 10020。

All rights reserved, including the right to reproduce this book or portions thereof in any form whatsoever. For information, address Scribner Subsidiary Rights Department, 1230 Avenue of the Americas, New York, NY 10020.

2022 年 10 月第一版 Scribner 精装版

First Scribner hardcover edition October 2022

SCRIBNER 和设计是 The Gale Group, Inc. 的注册商标,由本作品的出版商 Simon & Schuster, Inc. 授权使用。

SCRIBNER and design are registered trademarks of The Gale Group, Inc., used under license by Simon & Schuster, Inc., the publisher of this work.

有关批量购买的特别折扣的信息,请致电 1-866-506-1949 或发送电子邮件至business@simonandschuster.com联系 Simon & Schuster 特殊销售部。

For information about special discounts for bulk purchases, please contact Simon & Schuster Special Sales at 1-866-506-1949 or business@simonandschuster.com.

Simon & Schuster Speakers Bureau 可以将作者带到您的现场活动中。如需更多信息或预订活动,请致电 1-866-248-3049 联系 Simon & Schuster Speakers Bureau 或访问我们的网站www.simonspeakers.com

The Simon & Schuster Speakers Bureau can bring authors to your live event. For more information or to book an event, contact the Simon & Schuster Speakers Bureau at 1-866-248-3049 or visit our website at www.simonspeakers.com.

凯尔·卡贝尔的室内设计

Interior design by Kyle Kabel

Michael Nagin 设计的夹克

Jacket design by Michael Nagin

Yevgen Romanenko/Getty Images 的夹克艺术

Jacket art by Yevgen Romanenko/Getty Images

已申请美国国会图书馆出版数据编目。

Library of Congress Cataloging-in-Publication Data has been applied for.

国际标准书号 978-1-9821-7200-8

ISBN 978-1-9821-7200-8

ISBN 978-1-9821-7202-2(电子书)

ISBN 978-1-9821-7202-2 (ebook)

Original text